I have been monitoring news on the 2026 Ebola outbreak in the Democratic Republic of Congo, which in early August had more than 4000 cases and 1800 deaths. This makes it the second largest Ebola outbreak with little evidence of coming under control. It is caused by Bundibugyo Ebola virus, genetically distinct from those that have caused most earlier Ebola outbreaks. The Journal of Virology hosted a very informative webinar on this outbreak on August 12, 2026 which you can access here.

Although the reservoir for this specific Bundibugyo virus is as yet not firmly known, fruit bats of the family Pteropodidae are primary suspects. This reminded me that at the time of the COVID-19 pandemic, I had spent a couple of days looking into the ecology of bats and their viruses as the animals are important reservoirs of coronaviruses. I revisited that topic, as I found several intersecting aspects of bat evolution & behavior, virus replication & evolution, and anthropogenic habitat changes. The bottom line is that bats can serve as important reservoirs for viruses to propagate, evolve and migrate in ways that lead to ‘species jumps’ and outbreaks in humans. It is reasonable to posit this as a ‘complex system’ rich in causal mechanisms such that even though the probability of species jumps is very low, outbreaks of emerging pathogens become inevitable. Novel ‘emergent’ infectious diseases (EID) were rare events until the mid-20th Century but have been recorded at substantial frequency over the last 80 years, with more than half of zoonotic origin.

Bats are the second largest order of mammals, with over 1400 species, and serve as reservoirs for many zoonotic viruses, including filoviruses (such as Ebola), paramyxoviruses (such as Hendra and Nipah), and coronaviruses (such as SARS, MERS and SARS-CoV-2). Although several of these infections have a high mortality rate in humans, bats infected with them are asymptomatic. The underlying mechanisms for this are a good place to begin.

The dilemma of flight

A canonical response in most mammals to mitochondrial DNA present in cytoplasm (as might occur from cell damage due to bacterial or viral infection) is the synthesis of interferons and cytokines as well as initiation of programmed cell death (pyroptosis). The sustained powered flight of bats results in excessive cellular production of reactive oxygen species (ROS) and the release of DNA into the cellular cytoplasm from damaged mitochondria. But this inflammatory response would impede flight. The evolutionary response has been mutations that down-regulate interferon production and reduce the signals for pyroptosis in response to cytosolic DNA.

Viral infections frequently cause mitochondrial stress that leaks DNA into the cellular cytosol, and in most mammals (such as Homo sapiens) this leads to the symptoms of viral infection such as apoptosis or systemic inflammation. But the dampening of these responses in bats allows viruses to replicate while the animal remains asymptomatic. However, it is important to note that bats retain non-inflammatory functions of the innate immune response (such as alpha-interferon and interferon-stimulated genes that restrict virus replication) and that these are constitutively expressed.

The combination of dampened acute highly inflammatory responses with constitutively expressed, non-inflammatory virus-restriction factors creates a unique environment for virus evolution. Although virions are replicating, they are also subjected to constant, sublethal antiviral suppression. The consequence is positive selection for sophisticated immune-evasion strategies. If the virus can then spill over to another mammalian host that possesses highly inflammatory interferon and cytokine pathways, it can trigger cytokine storms, which in viruses such as Ebola, SARS and MERS lead to quite high case mortality rates in humans.

[Birds of course face a similar problem to bats. Their solution has been different: many of the genes involved in the mammalian inflammatory response have been completely lost. In addition, bird mitochondria generate ROS at a much lower rate and more efficiently repair their mitochondrial DNA]

Some basics of RNA virus genome replication

The zoonotic viruses found in bats are RNA viruses. The RNA-dependent RNA polymerases that replicate the genome of an RNA virus typically lack proofreading; the error frequency of ~5 x 10^-4 substitutions/site limits genome size to ~15 kB. Coronaviruses have a proofreading exoribonuclease that increases fidelity to 10^-6 substitutions/site and have the largest RNA virus genomes (27-32 kB). Note that this coronavirus genome error rate is at least 1,000 times greater than found for DNA replication. Hence, one expects that 10% of the 1-10,000 virions produced during a replication cycle will have a single nucleotide polymorphism somewhere within the genome.

It is also important to take note of the milieu in which this RNA virus is replicating – a bat cell that may be experiencing high levels of ROS (damaging to nucleic acids), and the constitutive expression of Interferon-stimulated genes that produce proteins which degrade genomic RNA or viral intermediates of messenger RNA as well as the ingress or egress of lipid-enveloped virus particles. One consequence of this highly reactive environment is that about 100 copies of a coronavirus genome are produced for every one incorporated into a virus particle.

A consequence of these constitutive anti-viral mechanisms is that bats can suppress but often do not eliminate virus infections. If the physiological status of the animal is stressed (for example, by nutritional starvation or birthing pups), suppression mechanisms may fail and animals will shed many more virions. The primary site of coronavirus replication is the epithelial lining of the intestines and thus viruses are released in feces (guano).

Bat roosting behavior as an incubator for virus evolution

The dynamics of evolution in these zoonotic viruses can only be understood in the context of bat roosting behavior. Horseshoe bats (genus Rhinolophus; a reservoir of coronoaviruses) inhabit karst limestone caves that run in a band stretching from southern China down through the Indochinese Peninsula. These caves often host multiple bat species, with population sizes that can range up to tens of thousands, particularly during the synchronous birth of juveniles during the wet season. The karst caves are deep, humid, and isolated from UV light. As virion-laden guano rains down it dries and becomes aerosolized; this generates a transmission mechanism whereby novel viruses can infect bats of either the same species or different species.

It is certain from analysis of anal swabs that individual bats harbor multiple viruses. What is unknown is the probability is that an individual bat enterocyte has become infected by more than one virus. Remember that bats suppress but do not eliminate virus intracellular replication. Hence, all that may be required is another infection via aerosolized guano. There is anecdotal evidence – most spectacularly the detection in rectal swab samples of Rousettus leschenaulti a novel coronavirus whose sequence showed had formed a chimera with an orthoreovirus gene. It is difficult to imagine how this arose other than from within-cell recombination.

The formation of viral genome chimeras is feasible because of the template-switching property of the viruses’ RNA-dependent RNA polymerases. After intitiation of replication, the protein may detach from the template but retain the nascent genome and re-attach to another RNA molecule in the cytosol. In this way, novel combinations of virus genes can arise.

The risk for novel virus generation is greatest in the weeks after birthing of pups. Females are metabolically stressed (therefore prone to shed more virus). As pups are weaned, the passive immunity provided by mother’s milk fades. There now is a large cohort of immunologically-naïve hosts that are exposed to a diverse assemblage of viruses from bat guano and these animals push viral concentrations even higher.

In this environment, single-site mutations may be less important than genetic recombination events that result in functional domain swapping to generate viruses with altered receptor-binding domains (that might include mammals other than bats) or that counter host’s defense mechanisms in novel ways (that might produce violent immunological symptoms in those other mammals).

Unintended effects of human activity

A fundamental epidemiological principle is that proximity of susceptible and infected individuals is directly related to pathogen transmission rates. A fundamental principle of bat physiology is that environmental stressors disrupt the balanced innate immunity mechanism and thus cause virus shedding.

Land-use changes such as deforestation impact the dynamics for tree-roosting fruit bats that are reservoirs for Nipah, Hendra, and Ebola viruses. The spatial structure of roosts changes when tree stands change from “dense” (many trees, fewer bats per tree) to “sparse” (fewer trees, higher bat density). Ecological modeling has demonstrated that sparse tree strands with high bat populations in the roost generate a greater force of infection, leading to larger, faster epidemic curves and volatile boom-and-bust cycles of viral shedding.

Deforestation also pushes fruit bats closer to humans and domestic animals as they access cultivated orchards to supplant lost native fruits; but this shift causes physiological and nutritional stress that increases viral shedding. A dormant reservoir may thereby be converted to a high-level emitter just when the bats are foraging in closest proximity to human and livestock populations. The genetic barriers for direct species jumps to humans are high, which is why intermediary domestic or agricultural animals are important. Commercial orchards, pig farms, and live-animal markets bring stressed bats into unnatural proximity with amplifying hosts. These intermediate hosts are an animal background where virus can adapt to non-bat mammalian cells and from there perhaps become capable of human-to-human transmission.

The dynamics for insectivorous horseshoe bats (Rhinolophus spp.) differ from fruit bats, but spillover risks remain. The pressures from horseshoe bats revolve around their adaptable nature and prey availability. Horseshoe bats are ecological generalists that can thrive in edge habitats created by forest fragmentation, for example, abutting agricultural lands. They may forage for insects around livestock operations putting them in close proximity to intermediate hosts who may become infected via guano drops. Thus, land use change does not just destroy horseshoe bat habitat; it generates “spillover zones” as generalist bats move into livestock areas. As habitats change due to anthropogenic land use changes or via precipitation and temperature shifts that alter arthropod emergence patterns, two causal factors may be enhanced. Expending energy to stay warm or hunt scarce prey unbalances the innate immune mechanisms that suppress viral replication; hence, more virus is shed. These habitat changes may result in migration shifts to caves whereby new combinations of bat species congregate in shared refuges with the potential for novel recombinant viral evolution.

Leave a comment

Coda

I really enjoyed delving into this topic. There are causal mechanisms operating from the subcellular to landscape scales that impact the evolution of these viruses. I believe there is scope to investigate these causal mechanisms more deeply, to unravel consequences of human behavior that may increase risks of virus range spillover to humans or outbreaks of human-adapted viruses. But this must also be done cautiously and responsibly. Just today I came across this op/ed by Alina Chan. She certainly has her own perspective, but it is undoubtedly true that analyzing these viruses with zoonotic potential ‘from the wild’ should be done under appropriate biosafety levels (BSL). But what is ‘appropriate’? There is a tendency to low-ball BSL as the work gets harder to do as BSL level increases. The number of BSL-4 labs has proliferated from 13 to 52 over the past 25 years. They (like biological field stations) can be at the ends of administrative and funding pipelines – when budgets get squeezed it is easier to defer structural maintenance and reduce staff training at this far end of the pipe, thereby increasing release risks. In my personal opinion, there needs to be a very high bar to justify experiments that might result in the generation of ‘humanized’ viruses (Gain-of-function research).

Today’s Moment of Zen

I have been working the last week or so to refresh my in-house picture ‘galleries’ that display prints of my photography. I now have an Asia gallery and a section devoted to black-and-white prints of architectural images. This post’s image is of the Great Buddha Dordenma, a 54-meter tall bronze statue of Shakyamuni Buddha located in Thimphu, Bhutan.

Including this is apt, as both my photography and working on this blog aid me in 2 aspects of the Buddha’s Noble Eightfold Path: Right Understanding (the true nature of things) & Right Mindfulness (paying close attention to my body and mind).

Subscribe now

After writing my last blog post on AI in biology, I explored the utility AI might have for the type of think pieces I do for this blog. As an emeritus professor at Purdue, I found that I could access Google AI Professional Certificate courses on Coursera, and they included a three month trial license for Gemini Pro (Google’s Tier 3 of 4). I went through 4 of the 7 on offer:

Words of Caution

I am going to put my summary thoughts here up front, for those not interested in my details on designing prompts. Doing those four courses was a reasonably short and painless way of gaining very rudimentary knowledge of using AI. The most direct question for me (as a commentator on the philosophy of microbial ecology) is its utility. Apart from summarizing research topics, there are capacities for searching for specific datasets as well as ‘data cleaning’ to free me from the tedium of tidying up and organizing Excel columns. I can get it to generate widgets/apps in a few minutes via ‘vibe-coding’ that can provide dashboards/graphical displays of data tables, and even play ‘what-if’ analyses via sliders that varied parameters.

In terms of summarizing areas of research, I remain extremely cautious. I constructed prompts for a couple of research topics. The results (using Gemini Pro Deep Research) were not bad but … I feel like even this relatively sophisticated AI remains a bit of a bullshitter (sensu Frankfurt).

I didn’t have to probe very deeply to find errors in citations, evaluations, or out-and-out fabrications. It helped to ask the AI to critically assess the strength of experimental evidence for its assertions. The main merits for me were for providing an initial overview that could lead to deeper investigations of topics I knew little about (e.g., I am thinking about a post on the ecology of bats and their viruses). I also find that AI may hit upon a different facet of a topic I thought I knew. That then stimulates me to search databases like Web of Science or Scopus for additional literature.

For those of you doing more than philosophizing, I don’t see that AI at the access level I used are useful for developing insightful new hypotheses. When I ask Gemini Pro Deep Research to do this, it ‘finds’ things proposed in scientific publications and regurgitates them. I mentioned Deep Mind’s CoScientist in my last post. It is very important to remember that whereas it took CoScientist days to arrive at its hypotheses, Gemini Pro Deep Research is only taking a few minutes for the task; hence these are very different animals. To end on a more positive AI-note, I saw this the other day:
“Nobel Prize-winning physicist and team use Claude AI to solve decades-old math puzzle” There is a link in the article to the scientists’ paper – way too many integrals, ODE’s and PDE’s for me …

Leave a comment

Prompting

Much of the material in the Google courses was geared towards business and marketing. However, there were elements applicable to science. The key is the prompts you provide it. Prompting is a skill that can be developed with practice and experience, but there are principles that are very useful in developing that skill.

Here are some higher-level principles:

• Be concise: A short, pointed prompt generally requires less processing from AI.

• Be explicit with precise and specific instructions.

• Be consistent: Use the same vocabulary for a concept throughout. Don’t use both ‘microbial community’ and ‘microbiome’ in a chat.

• Use follow-up prompts to correct assertions whose accuracy you question: Ask for confirmation, citations of underlying data and/or ask AI to assess the strength of the assertion.

• Be specific about the output: Add context about audience (microbiology undergraduates / microbiology graduate students / scientific experts in the field) and ‘product’ (a research report, a set of slides for a presentation, graphics, data tables) as well as the ‘persona’ you ask the AI to take on.

• Keep chats focused by starting a new conversation for each specific task. AI can refer back to previous information within a single chat and this helps keep its answers consistent. But if your prompts move to a different topic, this back-referencing can confuse issues.

• Save your best prompts in a prompt library

  Leave a comment

Note: I remain a novice at prompting. Major portions of this discussion have been abstracted from the Google AI courses.

First phase

• Define the AI’s persona by specifying the expertise or perspective it should adopt
  “You are a research scientist with a distinguished record in deep subsurface microbiology.”

• Clearly state the task you want AI to perform and specify the desired output format (a research report, a table, a graphic).
  “I am planning to write a review article on microbial activity in deep subsurface sediments. Research this topic for me and provide a research report with literature citations”

• Include relevant background information so AI understands what you want it to produce. You can upload pdf’s, images, or websites that act not only as a starting basis for AI but also as examples of style, tone, and format. When including references, make sure to explain how they relate to the task.

Evaluate the output

• Accuracy: Is the information correct and factually sound?

• Bias: Does the output give too much weight to one particular perspective?

• Clarification: Ask the AI whether it has questions about the task. Your answers may help it generate an output that meets your expectations.

• Relevancy: Did it answer your prompt and stay on topic you wanted? [if not, pursue refinements …]

Refinement

There are a number of reasons why the first prompt should not be the last one:

• The initial prompt was not clear enough -- add more details or clarify the persona, task, format, or context.

• Provide feedback or additional information and ask AI to update its response.

• If the response was too broad, add constraints to narrow the focus.
  “Focus the research on metabolic strategies to deal with extreme energy limitation”

• Rephrase the prompt to ask the same question but from a different framing.

• Was the prompt asking AI to solve a complex set of discrete tasks? If so, break into smaller parts and ask for step-by-step responses.

Advanced techniques:

1. Powerful prompt phrases: specific, precise language inside a single prompt to improve the quality and depth of AI’s response.

2. Prompt chaining: a series of smaller, connected prompts to structure the conversation and break down a large project into chunks manageable by AI

Powerful prompt phrases

These provide a pointer to how AI should produce a satisfactory output – they add specificity, set boundaries, and force it to use more complex reasoning paths.

Strategies:

• Give AI a process to follow

  • Think step-by-step.

  • First, [action 1]. Second, [action 2]. Finally, [action 3].

• Define the audience and tone

• Set hard constraints

  • Write a one paragraph summary.

  • Focus exclusively on [topic A e.g. physiology] and do not mention [topic B, e.g., applications].

• Request a critique (in Refinement stage)

  • How strong is the experimental evidence for this assertion?

  • Critique this text from the perspective of a bioengineer.

  • What is the strongest counterargument to this?

• Generate alternatives

  • Give me 3 different versions of this.

  • What’s an alternative approach to solving this problem?

• Deepen or expand on initial output

  • Craft a refinement that asks for more detail and/or specific examples for a particular section

Rather than provide my own examples (given my status as an AI novitiate), I asked Gemini Pro to develop Powerful Prompt phrases that could be used to generate a research report on biofilms. Here was the response when I asked that the prompt contain 3 action items:

Act as a distinguished Professor of Microbiology specializing in bacterial physiology and systems biology. Your task is to author a comprehensive, graduate-level research report on the mechanics and internal dynamics of bacterial biofilms.

Maintain an authoritative, rigorous, yet highly structured academic tone. Use precise scientific terminology and format the report with clear headings.

To construct this report, you must follow this specific process:

• First, detail the fundamental mechanics and lifecycle of biofilm formation. Break down the biological timeline from initial reversible attachment to irreversible adhesion, extracellular polymeric substance (EPS) matrix production, maturation, and eventual dispersion.

• Second, analyze the systems-level emergent properties and physiological shifts within the mature biofilm. Explain the spatial organization, metabolic stratification, and localized microenvironments that develop. Discuss how transcriptomic profiles and gene expression differentiate these sessile communities from their planktonic counterparts, using model organisms like Escherichia coli to illustrate these complex regulatory networks.

• Finally, evaluate current analytical frameworks and propose future research directions. Discuss cutting-edge methodologies used to study biofilm physiology (such as single-cell transcriptomics or spatial metabolomics) and synthesize these to suggest two novel, theoretical approaches for mapping biofilm dynamics.

Here was the AI’s explanation as to why this prompt is effective:

• Persona: Rather than a generic Microbiology Professor, the specialty is defined as “bacterial physiology and systems biology” – this primes the AI to focus on fundamental microbiological mechanisms rather than focusing on applied topics.

• The “Second” Action (Internal Dynamics): Requesting details on “emergent properties,” “metabolic stratification,” and specific model organisms is a guide for a technical analysis of community function.

• The “Finally” Action (Methodology & Synthesis): The third step now logically follows the second by asking the AI to evaluate the advanced research tools needed to measure these complex internal systems, resulting in a cohesive, basic science research report.

Prompt Chaining

An alternative approach is prompt chaining. This is particularly useful for complex tasks. Here, you plan a series of smaller (but connected) steps. The output from one prompt is part of the input for the next one. The result is a multi-prompt workflow that generates your AI solution. Prompt chaining may yield higher quality results for complex academic writing -- it forces the AI to focus its context window (“short term memory”) on one specific task at a time.

Prompt chaining would break up the biofilm prompt into a sequence of 4:

1. Persona + Action 1

2. Action 2

3. Action 3

4. Recheck the entire document to ensure sections flow logically.

Final Thoughts

An Information Technology staff member at Purdue told me the university hopes to roll out Gemini Enterprise-grade licenses to interested faculty and staff by the beginning of the fall semester (and may also do so for Microsoft 365 CoPilot; this premium tier integrates into Microsoft apps like Word and Excel). If my emeritus status gives me eligibility, I will certainly continue to use these tools (despite my caveats). I can think of several ways they would reduce the drudgery of some tasks. I don’t feel that using them to summarize scientific papers will be a best use for me — reading the papers and finding nuggets of wisdom therein is a delight.

Many hundreds of institutions seem to be forming these partnerships with Google or Microsoft — I encourage you to check with Information Technology regarding access.

Leave a comment

Today’s Moment of Zen

The World Cup came to an end yesterday — the final was a tactical, grueling match with little free-flowing artistry. But that was not an anomaly (and I knew that going in) — of the last 10 World Cup finals, 5 were 0-0 or 1-0 at the end of regulation time.

This was the second World Cup in the United States — I did not go to any matches this time, but had the TV on for most of the matches. The event embodied many of the contradictions present in my country and FIFA president Infantino certainly ‘Americanized’ some aspects of the event that did violence to the internal logic of the sport.

Congratulations to Spain for winning! I dug into my photo archive to find an image I had taken at the Mezquita located in Córdoba

.

The symposium was populated by a number of enthusiastic and highly competent scientists. Much of Hershberg’s summary was not on topics of particular relevance to microbial ecology. However, I did note his report on Pushmeet Kohli’s talk. I can do no better than to quote from Hershberg’s blog post:

“The last third of his talk focused on DeepMind’s efforts to build an AI Co-Scientist, which they published a new paper on last month. Given a research goal by a human scientist, this system develops a research plan and then kicks off a “tournament” of agents competing to develop new hypotheses. Agents within this system have different tasks. Some are designed to “reflect” on the ideas being generated. Others are tasked with “evolving” them.”

There is at least one practical example of how this worked. DeepMind partnered with Jose Penades, a microbiologist who works on capsid-forming phage-inducible chromosomal islands (cf-PICIs). The lab knew that helper phages were important not only to induce new particles but also to provide the phage tails necessary to infect other cells. What puzzled them was that rather than a narrow host range, cf-PICIs were found in a broad range of taxa – from Gammaproteobacteria to Bacillus sp. to perhaps Actinobacteria. They carried out a series of experiments and concluded that cf-PICIs released tail-less capsids but used tail adaptor and connector proteins to interact with phage tails released from other cells (and including those from other bacterial species) in the extracellular milieu.

What was crucial for a test of CoScientist was that the Penades group had submitted a manuscript on this work but that information was not in the public domain (and hence not accessible by an AI agent). From a video interview of Penades at the CSHL symposium: he was excited because

“not just that the top hypothesis that they provided was the right one, it was that they provided another four and all of them made sense. For one of them, we never thought about it, and we are now working on that.”

Penades et al. (2025) published the results of this ‘experiment’ here:
AI mirrors experimental science to uncover a mechanism of gene transfer crucial to bacterial evolution

In it, they document the preliminary information provided to the agent (all of which was in prior public domain) and the research directions the agent developed. Here is their table of 6 proposed research directions and a brief idea of underlying hypotheses:

As noted, the top research direction suggested by CoScientist was experimentally shown to be the correct one. But Penades thought that some of the alternative ideas were creative and most significantly, testable. Some were less novel (see New Areas of Research column in table), but others have stimulated their thinking to pursue new directions for satellite biology.

They provided the same information to 7 other agents (e.g., Gemini 2.0 Pro Experimental, OpenAI deep research, Claude Sonnet 3.7 and DeepSeek-R1) and none were able to generate ideas regarding capsid-tail interactions. Hence, AI-CoScientist has some particular properties [however, note that the computation process requires several days rather than minutes for the other agents].

I thought that Penades et al., as experimental scientists, made some important points regarding the future use of AI agents:

1. If the ‘conventional wisdom’ within a specific subdiscipline is limiting humans’ creativity, AI agents may identify novel approaches. I loved doing multidisciplinary research for this reason – scientists from other disciplines viewed different facets of a problem than I did. Novelty could arise from integrating disparate views and approaches.

2. These AI agents may very well generate novel ideas and many may be theoretically possible. However, they require critical oversight by expert scientists as the agents lack contextual judgment.

I worry a bit about the use of these agents for training the next generation of scientists. Will they be a crutch that inhibits development of ‘contextual judgment’ or an aid that permits more opportunities to consider context?

Leave a comment

Hypotheses

Penades et al. referred to CoScientist’s “hypotheses.” If I look at the output, they feel more like research directions. (Although this may be a consequence of how they reported the outputs; the specific ideas in the table above could be reformulated as formal hypotheses). This is a very key point for the value of these agents for future research. As Jim Prosser has pointed out much of published microbial ecology research is now hypothesis-free. In another important paper, he addressed the philosophy of doing experimental microbial ecology.

As I am not an expert in satellite biology, I cannot assess how deep or ‘risky’ (sensu Popper) are hypotheses CoScientist developed. That Penades finds them ‘testable’ (falsifiable) is certainly a mark in their favor.

For future uses of this tool, I think including a philosopher of science in the development process would be highly valuable. That individual could be tasked with ensuring that words in prompts are used precisely (see my post on Mechanism). Refining the output so that it produces testable, risky hypotheses rather than general research directions would be very valuable to experimental scientists who appear to rarely think deeply in such terms. In that fashion, an AI agent such as this can help meet Prosser’s challenge:

“be brave enough to look for interesting but maybe difficult questions, propose bold and risky hypotheses, test them critically and be prepared for them to fail. This, rather than relentless, aimless accumulation of data, has and will provide real advances in microbial ecology.”

Leave a comment

Today’s Moment of Zen

In between watching World Cup matches, I have enjoyed working on images I took on my Danube cruise through southeastern Europe.

This was an interesting shopping complex in Bratislava, Slovakia. The striking Monolith (Obchodný dom Prior) with clean, unyielding geometric lines was designed by Ivan Matušík and built in the 1960s. The panels of Slovakian Spiš travertine contrast with the bold lines of the structure and the embedded marble clock provides a central focal point.

Prefatory remarks: The opinions arise from the philosophical basis for Think Like a Microbe -- that microbial ecology is best served by research that identifies principles and their underlying mechanisms.

I decided to determine which habitats were being studied by microbial ecologists. I used a Web of Science search for published research on microbial ecology over the past 10 years. That returned about 60,000 articles. If I restricted the search by a topic such as the habitat “soil” or “ocean,” here are the subsets that were returned:

Leave a comment

About one-third of the publications concern soil, although this is probably the most difficult system to interrogate at the (small) spatial scales where environmental and microbial characteristics exhibit heterogeneity. Oceans come second – I was expecting a much larger difference between them and freshwater habitats (lakes and streams). I understand why oceans are important, but they also represent a system where the logistics and expense of conducting research are so much greater than on lakes whereas the ecology of planktonic bacteria is nearly identical (Hobbie,1988).

Studies on the gastrointestinal microbiome rank third. Genomic analyses of fecal samples have produced an explosion in this field over the past 20 years:

As I mentioned in an earlier post, it remains an open question for me the extent to which end-member fecal samples of a plug-glow reactor are representative of the ‘action’ in the GI tract.

In my opinion, studies on the pathogenesis of infectious diseases are also ecological. These papers would not be captured by ‘ecology’ in a search term, so I did a separate search for ‘pathogenesis’ and ‘virulence’ which returned 44,000 articles – about 75% of the number found for all other habitats combined. If I expand the search to the topic ‘pathogen’ (although many of these would have no ecological context at all), there are 6X as many papers.

The science-of-microbial ecology science revisited.

I had a post on this topic in January, 2025 but provide a revised analysis here for readers who have come onto this more recently.

What countries are providing research funding to support this area of research? I used Web of Science’s analysis of cited funding agencies and included not only the 60,000 articles with ‘ecology’ as a hit but also the 44,000 for pathogenesis. Here is a list of the countries/regions that were most highly represented:

The countries of origin are relevant, as most research funding comes from national agencies. I cannot speak to the criteria upon which research projects are funded in any country other than the United States.

Within the U.S., the National Science Foundation is the organization whose mission is best aligned with fundamental research and the one where scientists on a panel have had the most direct influence on what specific grants get funded. When I served on the Ecology Panel, the panelists preferred grants that addressed a broader ecological theory; this made it difficult for microbial ecologists at that period of time. Microbial ecology could get funded through ad hoc programs such as LExEn (Life in Extreme Environments), started in 1997 but whose funding has now been incorporated into the core Directorates.

The National Institutes of Health, by far the largest funder of biological research, is of course “mission oriented.” However, the National Institutes of General Medical Sciences (NIGMS) was historically a very important source of basic research funding. The impetus was to understand fundamental mechanisms in cell biology and from 1960 through the 1990s, bacteria (particularly Escherichia coli) were the most tractable systems. Fundamental research remains funded here, and has shifted to include the study of microbiomes but relevance to human health remains important.

The Department of Agriculture (USDA) is explicitly mission oriented, and it is not surprising that the microbial ecology it funds is dominated by work in soils or upon plant pathogens. The Department of Energy (DOE) is a more interesting case – the mission of the Biological and Environmental Research program is to “achieve a predictive understanding of complex biological, earth, and environmental systems for energy and infrastructure security.” That can cover a broad territory and indeed they have funded work on array of terrestrial and aquatic habitats. Within DOE, program officers have more control over who gets funded than at NSF. When I served on review panels, it was made clear that we were there as “advisers” to the program officers. BER programs tend to have a finite lifetime (5-10 years) although they may morph into new programs that have a connection to the past. For example, Frank Wobber was a DOE program officer who championed the study of subsurface microbiology beginning in the 1980s with the Subsurface Science Program. In the 1990s, this transitioned to the Natural and Accelerated Bioremediation Research (NABIR) program that sat at the intersection of microbial ecology, biogeochemistry, and subsurface science, with an objective to examine contaminant fate and transport at DOE sites such as Hanford, Oak Ridge, and uranium mining sites. One *could* do fundamental science here as long as it was tied to understanding a contaminant of interest. NABIR connected microbes to real-world environmental processes at management-relevant scales. That work has continued – there has been a continuing push towards larger scale systems, which is needed but also at some variance with the spatial scale at which I think principles in microbial ecology are most likely to be discovered.

Real vs. tractable

In his 2002 Nobel Prize lecture, Sydney Brenner said “We’re drowning in a sea of data and starving for knowledge.” In a subsequent interview he said: “but there is now a crisis developing in biology that completely unstructured information does not enhance understanding. What people want is to understand, which means you must have a theoretical framework in which to embed this.” We are now more than 20 years down the road and the sea of data in microbial ecology has increased exponentially. In 2002, there were 100 microbial genome sequences available; as of June 2026 gcMeta contained nearly 4 million Metagenome-Assembled Genomes (MAGs) from about 128,000 environmental samples.

Certainly large sequence datasets collected from natural habitats in time or space can by used to infer the relative importance of deterministic vs stochastic processes in microbial community assembly. But the investigators generally fail to explore the detailed mechanisms of selection / cooperation / niche breadth / dispersal that determine those processes in planktonic / unsaturated soil / biofilm habitats.

Leave a comment

Prosser et al. (2007) argued 20 years ago that advances in microbial ecology were not limited by techniques but by inadequate conceptual approaches. They recounted which of the rich theories in macrobe ecology were applicable to microbial ecology (and indeed where microbial systems could rigorously test theories difficult to interrogate in plant or animal systems). There have been some attempts to do so but many important theories have been understudied – areas such as spatial and temporal heterogeneities driving activity and community composition and trait-based theory as it pertains to functional redundancy in microbial communities.

In 1988, Slobodkin said: “Ecology may be the most intractable legitimate science ever developed.” He then compared this intractability to late 1980s medicine. Yet we have seen incredible advances in medicine over the intervening 4 decades. Yes, technology has been important but that technology rested upon fundamental advances in mechanisms of biochemistry, cell biology and the immune system. For example, I have had recurring sinus problems for more than 40 years. I just started injecting monoclonal antibodies against a cytokine which is an upstream activator of Type-II inflammation. I can remember when monoclonal antibodies were first developed in mice as a research tool and now that understanding has translated into a therapeutic.

What is needed to make microbial ecology more “tractable?” Certainly theoretical advances are key, but so are approaches that privilege islands of knowledge over seas of data. I believe that a greater use of model systems / “reductionist” science will advance this goal. I addressed this a bit here . As I said then: “These studies certainly are not ‘real.’ But are experiments with natural samples ‘real’? Any sample is a snapshot of a time and specific place that may not be truly representative of the system dynamics going forward.”

Some characteristics of a useful research program are simplicity and the capacity to solve existing puzzles and support/refute risky hypotheses. I would argue that the disadvantage of a well-chosen model system’s ‘unreal’ nature is an advantage in that it allows the experimenter to make conscious decisions regarding which complexities to include and exclude. In this way, a model system interrogates patterns uncovered in the sea of data and thereby can address fundamental ecological questions to produce ‘knowledge.’ It is the high level of complexity in natural systems that makes ecology ‘intractable.’ Model systems / microcosms / mesocosms are indeed caricatures of nature. But they offer more promise for unraveling the mechanisms and principles of microbial ecology.

Am I optimistic that funding agencies will affirm and reward this fundamental approach? Not particularly, as ‘mission-oriented’ programs are more easily understood by politicians who appropriate taxpayer dollars for scientific research. It would be nice if a private foundation could support this type of effort, as the Moore Foundation and Simons Foundation have supported research on marine microbiology in recent decades (their support has been cited in about 500 papers on microbial ecology over the last 18 years). The header for this section is a paraphrase of a Led Zepplin: “What Is and What Should Never Be.” My comments here, like the lyrics of that song, are a romantic fantasy given the realities scientists face in maintaining and funding a research program.

Today’s Moment of Zen

Zaha Hadid is one of my favorite architects and when I was in Vienna last month I passed by the library on the Universität Wien campus that she designed. I had to go back later in the day to take some images.

Subscribe now

When I was researching material for my Systems Thinking post, I came across W Ford Doolittle’s book Darwinizing Gaia. Doolittle was an early critic of the Gaia hypothesis, on the basis that it viewed Earth as an organism upon which has evolved via natural selection. As variations arise during reproduction and the Earth does not reproduce, he found this mechanism untenable. In this latest book, he admits the possibility that ‘persistence’ in addition to differential reproduction could be a selection mechanism.

This got me thinking about the microbial world and the many examples whereby ‘persistence’ operates via a reduced death rate of propagules.

These include production of morphologically recognizable resistant forms:

1. Endospores formed by Firmicutes..

2. Akinetes formed by cyanobacteria in the orders Nostocales and Stigonematales. These dense cells can sink to the bottom of aquatic ecosystems and remain dormant until environmental conditions improve.

3. A significant number of fungal and actinobacterial genera form aerial mycelia that contain conidia which function not only for dispersal but also as metabolically quiescent cells.

4. Similarly, myxobacteria can form myxospores withing their fruiting bodies, and represent dormant propagules.

5. Azotobacter species form morphologically-distinguishable cell types called ‘cysts’ that have a reduced metabolic rate and are dessication resistant.

Endospore formation (particularly in Bacillus subtilis) has been studied for many decades as a model of differentiation The volume of literature on conditions leading to dormancy or resuscitation for categories 2-5 is very modest.

Other categories related to dormancy:

1. Viable But Nonculturable (VBNC) cells

2. Persister cells that are not killed by antibiotic treatments

3. Rod-to-coccus transition for several non-mycelial actinobacteria, such as Arthrobacter, Micrococcus and Rhodococcus. [Also note that the actinobacterium Mycobacterium tuberculosis persists in tubercules for years in a dormant state.]

4. Aeonophiles/Anergiobiosis Organisms with incredibly slow metabolic rates that persist in deep subsurface sediments, presumably over geological time scales.

With respect to #3, it feels fair to mention that all rod-shaped cells become smaller as growth rate slows and hence will approximate a spheroid shape. As its doubling times decrease from 120 to 20 minutes, the cell volume of Escherichia coli increases 6-fold. Why should that be? As about 50% of cell mass is protein, growing faster means synthesizing protein faster. Increasing the ribosome content of a cell is a facile way to do that --over that range of doubling times, the cellular ribosome content increases from about 7,000 to 70,000. Increasing the ribosome content of a cell is a facile way to do that and ribosomes take up space. That many ribosomes can only be ‘housed’ in a larger cell volume and hence the small volume of slow growing or dormant cells merely reflect growth rate control of ribosome synthesis.

Measuring dormancy in nature

There are about 250 scientific papers published per year that have dormancy as a topic (over the past 10 years); about one-third of these appear to have an ecological focus. Relatively few actually measure the size of the dormant population – many cite Lennon and Jones (2011) which contains data from the 1990s and 2000s, with a conventional wisdom that most soil bacteria are dormant.

The methods used were the best at the time but certainly have limitations. The most interesting current methods include the use of ¹⁸O-water to label nucleic acids (qSIP) and Bioorthogonal Non-canonical Amino Acid Tagging (BONCAT) to label cells actively synthesizing proteins. Harris et al. (2025) applied BONCAT to bulk soil, rhizosphere, plant roots and root nodules and obtained results consistent with conventional wisdom. However, Papp et al. (2018) found that >90% of soil taxa (not cells – important note) incorporated ¹⁸O-water into RNA and DNA during incubation periods of 1-8 days. These experiments all “suffer” from a disturbance effect –soils are amended and/or perturbed which is known to stimulate a flush of activity (measured as CO₂) evolution. NanoSIMS or Raman-SIP would be an excellent technology with which to combine the advantages of stable isotopes with individual cell measurements.

These latter technologies are much more robust than what was available 15 years ago. Many more analyses of the proportion of metabolically-inactive cells (and the dynamics with which they regain activity upon environmental upshifts) are warranted to move the field from a state of conventional wisdom to a broader analysis of in what habitats and under what circumstances ‘dormancy’ is quantitatively important.

The dynamics of going dormant.

For me, the ‘classic’ paper describing these phenomena was published 30 years ago and summarized a set of work in the Kjelleberg lab on a Vibrio sp.

Kjelleberg, S. et al. How do non-differentiating bacteria adapt to starvation?. Antonie van Leeuwenhoek 63, 333–341 (1993).

Their work identified several adaptations that have also been identified in a number of less comprehensive studies. Cell size decreases (and becomes spheroid). ppGpp is the stringent control response ‘alarmone’ with pleotropic affects. There are ‘starvation-induced’ proteins; in other studies some of these have been claimed to promote ‘hibernation’ by dimerizing ribosomes and slowing activity of proteins such as RNA polymerase. Note that some of these proteins (in the 0.5-6 h window) cast a wider net for alternative resources (such as an exoprotease or chemotaxis receptors).

Unfortunately, there have been few other studies in the intervening 30 years which have mimicked this analysis using approaches such as transcriptomics. The utility of such experiments would be to determine the extent to which there are general principles that could then be broadly applied or idiosyncratic responses by microbes from particular habitats or evolutionary history.

Yes or no … the game theory of dormancy

Application of the Modern Evolutionary Synthesis posits that ‘dormancy’ is selected because it increases fitness of the organism over evolutionary time scales. But certainly environmental and physiological factors must impact whether fitness improves by going dormant. Some of these include:

How frequently (and regularly) do cells experience starvation, dessication, etc.? For cells in the human GI tract, resources arrive several times a day. Diurnal and tidal rhythms are relatively short term and quite regular. Microbes in a desert soil may experience long periods of aridity, punctuated by brief, intense rainstorms. The aeonophiles apparently persist over millenia without substantial nutrient resources.

Lennon et al. (2021) covered several aspects of theory and how mathematical modeling approaches can be valuable to sort out the environmental and biological dynamics under which transitions to dormancy might have a selective advantage.

Crucial questions include:

1. How much of a decrease in death rate accrues from the dormant state?

2. How frequently, regularly and for how long do adverse conditions arise?

3. How long does resuscitation take after conditions improve?

In a very simple simulation model, Konopka (2000) found that the selection coefficient was most sensitive to the time constant for conversion of dormant cells back to actively growing ones, and the magnitude of the death rate of cells which did not undergo the starvation response

Resuscitation from dormancy

For more than 40 years, microbiologists have been aware that cells can enter a “Viable But Non-Culturable” state and sought means to resuscitate them. James D. Oliver has done extensive work on this topic. Of particular interest are taxa that may be human pathogens, as their survival as VBNC in nature may impact disease transmission or reemergence. Of particular note are intragranulomatous tubercules that are believed to harbor dormant Mycobacterium tuberculosis. Estimates of latent tuberculosis infection range from 20-30% of the world population.

Resuscitation conditions may differ among taxa and include procedures such as temperature shifts. Notably, a number of actinobacteria have been found to generate “Resuscitation promoting factors” and genes have been identified. These appear to be enzymes that enzymes that remodel the peptidoglycan cell wall, and generate fragments that may stimulate other cells to resuscitate. This observation is consistent with Epstein’s scout hypothesis of a few active cells sampling the environment and if there is a growth response, the resulting muropeptide fragments serve as a signal to nearby dormant cells to resuscitate.

As with the dynamics of becoming dormant, transcriptomics is a powerful method to analyze resuscitation. In a study that employed ²H₂O as an activity measure in a simulated rain event upon Negev Desert soil, Imminger et al. found that resuscitation was rapid (~30 minutes) and >90% of cells regained activity. Via metatranscriptomics, they were able to identify taxa and specific gene classes that were up-regulated.

Friend-of-the substack Paul Carini is interested in the aridity experienced by microbes in desert soils. He and his colleagues recently published a very interesting paper in which they analyzed the transcriptional responses of an Arthrobacter strain to first dessication, and then rehydration by exposure to *water vapor* rather than water. They identified 6 co-expression modules during these transitions:

Don Cowan provided a very nice commentary on this article

https://journals.asm.org/doi/10.1128/msystems.00493-25

and some more general comments on the microbial ecology in hyper-arid soils.

In summary, dormancy is an extremely important topic in microbial ecology particularly in certain habitats (unsaturated soils, hyper-arid soils, deep subsurface sediments, and the lesions of those who have been infected by Mycobacterium tuberculosis). However our mechanistic understanding (in the sense discussed here remains very rudimentary, despite the technological capacities to interrogate this phenomenon under ecologically-relevant conditions. I mentioned early in this article that about 80 papers are published per year on ecological aspects of microbial dormancy – that is <1% of all papers published on microbial ecology in 2025.

Today’s Moment of Zen

This is my first posting since returning from a two week trip up the Danube River, starting near Bucharest, Romania and ending near Munich, Germany. This was a great trip in several ways, including my interest in architectural photography. There were a wide variety of architectural styles on offer in cities like Bucharest, Belgrade, Budapest and Vienna. These included Gothic, Renaissance, Baroque, Neo-Classical, Art Noveau, Vienna Secession, Modern, Brutalist, and Deconstruvist. Wow!

Here is a blue-period view of the Hungarian Parliament in Budapest, taken as we sailed upriver toward Bratislava.

Leave a comment

Subscribe now

If I were given the challenge to design the curriculum for an introductory biology course, I would include a couple of lectures on Systems Thinking. This would be less abstract than the biologist Ludwig Von Bertalanffy’s General Systems Theory of the late 1960s -- it would encapsulate “models, principles, and laws that apply to generalized systems or their subclasses, irrespective of their particular kind, the nature of their component elements, and the relationships or “forces” between them.” Nor is it synonymous with what is recognized as systems biology by molecular biologists – the generation of large transcriptomic datasets and their subsequent analysis to identify patterns or networks. O’Malley and Dupre (2005) described this as ‘pragmatic’ systems biology. Mine would be their ‘systems-theoretic’ biology with an intent to identify principles that are broadly applicable to different specific biological systems and (perhaps) at multiple hierarchal levels. As I started to think about this, I realized it dovetails very well with Craver’s depiction of a mechanism underlying a phenomenon, that I discussed in this post.

If you recall, Craver and Darden proposed that experimentalists devise multi-level schema of how system components interact with each other to better explain how a phenomenon is caused. That is systems thinking in a nutshell. It is both art and science – making predictions about systems behavior by developing an increasingly deep understanding of underlying structure.

My intention for adding this to a biology curriculum would be to expose students to a new/different way of thinking about biology. Not to stop at a collection of trees (e.g., regulation of tryptophan biosynthesis in a cell) but think about the network that regulates the rates of biosynthesis of all 20 amino acids such that each are in balance with each other and with the cell’s demands for protein synthesis.

What are some fundamentals?

Vantage point is one essential fundamental. Per Craver and Darden’s concept of schema consisting of a nested hierarchy, systems thinking requires detailed analyses of sets of mechanisms at multiple levels to understand overall system dynamics

Some important principles include:

1. It is the set of interacting components in a system (and their responses to internal or external perturbations) that determines system behavior, rather than external forces per se.

2. Closed feedback loops are a fundamental structure regulating system behavior.

3. Many subsystems can be visualized as a series of stocks/pools that have levels/concentrations with flows both into and out of them. It is the infrastructure of stocks and flows that provide the basis for feedback loops that control the system.

SIR epidemiological models (with stocks of Susceptible, Infected, and Resistant individuals) were first developed 100 years ago by Kermak and McKendrick (and remained a mainstay of epidemiological modeling at the inception of the COVID-19 pandemic!). Here is a graphical example of a stock & flow SIR model developed in 2020 using the simulation software Vensim.

You can see the basic structure in the 5 black Stocks, with flows between them. The flows are calculated as simple algebraic expressions, using values from elements with arrows pointing to the flows. For example the flow Becoming Symptomatic is calculated as Infected asympto/Incubation period. The modeler sets a time step over which flows and stocks are updated, effectively integrating dynamics over time. Homer has added a number of other factors whose values can be adjusted and their effects on the basic system evaluated. [Note: epidemic modeling over past 5 years has moved significantly towards individual agent-based models that incorporate stochasticity and temporal & spatial heterogeneities]

Simulation modeling

Thinking about systems and their component interactions raises consciousness regarding all strata of biological organization. Taking the next step with modeling simulators as simple as Vensim can provide predictions of short- and long-term consequences of altering components upon the system’s dynamics. Interfacing such modeling with experimentation to test model predictions can deepen our understanding of mechanisms.

As one example, I constructed a simple model of this type to compare the fitness of a bacterium that shifted to a dormant state under nutrient starvation to one that did not. I varied the time interval between nutrient pulses and did sensitivity analyses in which the organisms’ kinetic parameters were systematically changed under the hypothesis that longer intervals would favor dormancy. The conclusion was that the selection coefficient for dormancy was most sensitive to the time constant for conversion of dormant cells back to actively growing ones, and the magnitude of the death rate of cells which did not undergo the starvation response. Subsequently, a Master’s student, Militza Carrero-Colón tested these predictions. Sadly, they were not supported – but it does represent a case of how even simple simulation modeling can inform subsequent experimentation

I also used simple graphically oriented simulation models such as the SIR models in teaching some of my microbiology courses at Purdue. For purely undergraduate courses, I built the model in the background and presented the students with a dashboard whereby they could control the simulation parameters.

For advanced undergraduate/graduate student course, I gave students a primer on how to construct models in software like Vensim and had them do a class project of their own design.

I found that engaging students with dynamic modeling of this type had some important outcomes:

1. They could experiment/play with the models. They could ask ‘what-if’ questions by changing paramaters and observing how system behavior changes.

2. A simulation model functions as a thought-organizing device (per Craver & Darden’s schema). For students, the value arises from a deeper understanding of how variables impact system behavior that static lectures to them cannot deliver. For researchers, the detailed model/schema may demand assembly of experts from other disciplines who can delve into specific subsystem assemblies. It is a great means to foster interdisciplinary collaboration as members gain knowledge from each other at the disciplinary boundaries of the system.

On feedback – negative vs positive

Most feedback loops in systems are negative and act to modulate rates and/or dampen oscillations. However, there are examples of positive feedback. Bacterial growth is an obvious one – it is auto-catalytic. The product of growth is more units that can catalyze growth. Of course, in a real-world system the depletion of an essential growth nutrient is a negative feedback that slows or stops growth rate.

A concern in Earth climate science is the fate of organic C sequestered in permafrost at high northern latitudes. The amount (~1500 Gt) is approximately 40 times the amount of fossil fuels burned each year. The rise in atmospheric CO₂over the last 200 years has led to temperature increases at these high latitudes that have begun to melt permafrost, thereby making the organic C available for microbial metabolism. Current estimates are that ~10% of permafrost has thawed. If anoxic, then methane (another greenhouse gas) is produced, but once oxic, organic C will be mineralized to more CO₂. This will represent a positive feed-forward in which thawing of permafrost leads to more CO₂ release to the atmosphere which will drive an accelerated rate of warming and permafrost thawing.

Leave a comment

Complex systems

Once one embraces systems thinking, the concept of complex system emerges [pun intended]. In 2009, I spent three days in a Washington DC hotel at a workshop for the U.S. Department of Energy’s Office of Biological and Environmental Research on complex systems science. I was looking forward to this as I had been engaged in systems thinking for a decade or so. However, it felt like significant time each day was spent re-litigating what made a system ‘complex’ rather than merely ‘complicated.’

Complex systems are thought to have ‘emergent’ properties – the self-organized formation of collective structures and behaviors that are distinct from those of individual entities that comprise the system. In microbial ecology some examples of emergent properties are:

1. Biofilm Formation / Spatial Self‑Organization

2. Metabolic Cooperation (Cross‑Feeding)

3. Community Resilience to environmental perturbations

More generally in macrobe ecology, these types of phenomena are considered emergent properties:

1. Tipping points that trigger large regime shifts in the system

2. Trophic cascades, where the effects of a predator ripple down through a food web and may even impact the physical landscape.

3. Functional Redundancy and Response Diversity

4. Biogeochemical cycles of nutrients and energy

Defining or identifying what constitutes an emergent property can be fraught and contentious. In my view, a system is ‘complex’ if the relationships between system constituents are not strictly additive or linear. Nonlinear dynamics can result in even small perturbations producing a very large effect on system behavior and can lead to chaos (in a mathematical sense), periodic or nonperiodic oscillatory behavior, or the type of fractal organization seen in bacterial colonies.

Earth System Science

Lovelock and Margulis (1974) created quite a stir 50 years ago: “The notion of the biosphere as an active adaptive control system able to maintain the Earth in homeostasis we are calling the ‘Gaia’ hypothesis.” I think I was turned off by the mystical connotations at the time as well as the idea of Earth as a living organism. If we just call it Earth Systems Science and emphasize the physical, chemical and biological feedback loops that regulate processes I am OK with it. But there is incredible complexity that we very incompletely understand. Consider that despite the advances in physics models, satellite sensing coverage and computational power made in the last 50 years, predicting the weather 10 days from now is no better than a coin flip.

I came across a book written by the eminent microbial molecular and evolutionary biologist, W Ford Doolittle, Darwinizing Gaia. It is open access and I highly recommend it.. He includes very nice chapters that summarize the Gaia Hypothesis and the Modern (1940s) Synthesis of Evolution. Margulis extended Lovelock’s original Gaia hypothesis from a homeostatic mechanism to one that is autopoietic and evolutionary. Doolittle was a vocal Gaia naysayer – on the basis that Earth can’t be an ‘organism,’ because natural selection operates when organisms reproduce and Earth does not. However, in this book he is at least open to arguments for Gaia on the basis that selection *might* also operate not only via reproduction but also by persistence.

Summing up

I remain enthusiastic regarding all biologists engaging in systems thinking, developing schema of how they perceive the mechanisms that underlie their systems and even engaging in systems modeling of stocks and flows to improve their intuition about their dynamics. However, I also recognize that deeper steps require much more and if I look at the literature, progress seems rather slow. One impediment I articulated in an earlier post on Mathematical Modeling:

“Progress requires effective intellectual engagement between microbial ecologists and mathematicians. It would be rare for a single individual to function at the highest level in both ecology and mathematics.”

The nature of complex systems (i.e., those that contain emergent properties) is that they can exhibit nonlinear, unstable and aperiodic dynamics. Investigators may attribute time series with such properties as “noise” or measurement error, biased by the long-standing idea that populations and communities exhibit a “balance of nature.” But what if the actual dynamics in systems are chaotic, with instabilities and sensitivity to initial conditions? That requires a different set of mathematics with which even mathematically-inclined biologists are uncomfortable.

I have found George Sugihara’s work most interesting in this regard. Sugihara trained under the late, great Robert May and starting in 1990 investigated nonparametric mathematical methods of exploring time-series data to reconstruct ‘attractors’ that disentangle apparent ‘chaos.’ This approach most recently has been called Empirical Dynamic Modeling. In the end, this approach illuminates phenomena but the attractors may point to high-level mechanisms such as competition or predation.

Source: Tracking and forecasting ecosystem interactions in real time

Today’s Moment of Zen

We are leaving home in about a week for a two week cruise on the Danube River, from Romania up to southern Germany. This got me thinking about earlier trips to Eastern Europe, and I pulled this from my photo archive, taken on a one week road trip through Slovakia prior to the 2006 ISME meeting in Vienna.

Subscribe now

Several months ago, I watched a webinar sponsored by the U.S. National Academy of Sciences, “Was Science’s Social Contract Just a Myth?”.

I found it quite intriguing – it gave some background on the idea, a trajectory on how it has unfolded over the decades, and what it might mean going forward. Hence, I decided to dig a bit deeper.

History of the tacit ‘social contract’

The idea underlying a ‘social contract’ between primarily university scientists and U.S. society dates to the end of World War II. Vannevar Bush presented a report to US President Harry Truman in July, 1945 entitled “Science – the Endless Frontier.”

The report argued for federal agencies that would fund scientific research at universities. This led to rapid expansion in the existing National Institutes of Health and creation of the National Science Foundation in 1950. Bush’s argument was that if the country nurtured cutting-edge basic science, inevitably new technologies would arise with impact on economic growth and job creation, national security and improved health even in the absence of an explicit national industrial policy targeted to those ‘goods.’ The tacit social contract with scientists was that they would be provided with a large degree of self-governance and financial support with no strings attached for basic research. In the last 25 years, we have seen the development of pluripotent stem cells, mRNA vaccines, CRISPR gene editing, personalized medicine, therapeutic use of monoclonal antibodies etc. All of these technologies arose from fundamental biological science research.

In principle, this social contract in the United States applies to basic research funded by the federal government. The National Science Foundation tallies all Research & Development (R&D) – public and private -- that occurs every year. Here are the increases in federally-supported basic research over the last 75 years.

Several points about R&D in the United States

• Basic research historically was mostly undertaken by faculty at universities. However, now only about 50% occurs there and 30% is performed in the business sector.

• Funding for basic research has ranged from 0.1-0.5% of U.S. Gross Domestic Product over the past 50 years.

• Applied research monies have had a similar temporal trajectory but are about 120% of basic research expenditures each year.

• Research funds are dwarfed by Development (“systematic use of the knowledge or understanding gained from research, directed toward the production of useful materials, devices, systems, or methods, including design and development of prototypes and processes.”). Development comprises two-thirds of the $1 trillion R&D spending in 2024.

• About 75% of U.S. R&D takes place in the business sector – most of it in “Development”.

• U.S. federal spending on R&D is dominated by the Department of Defense at ~$185 billion. NIH is by far the next largest supporter of R&D (at $48 billion) but it and all other federal agencies are only 25% of DoD expenditures.

Anecdote #1: I began as an assistant professor at Purdue in 1977. My senior colleagues would regale me with stories from the 1960s how program officers from NIH would call them and beg to submit grant proposals so that they could fund them. Things were already different at the time I began. When I saw that my lab-neighbor Ed Umbarger, a National Academy of Sciences member, was sweating over renewal of his NIH grant, I knew that it would be a slog for me.

The panelists on the NAS webinar all agreed there have been large changes in the basic science enterprise and the public’s perception of it. I see two major changes. One is a diminution of ‘pure’ basic science for mission-oriented funding programs not only in agencies like the Department of Energy and U.S. Department of Agriculture but also in NSF.

Anecdote #2: I knew a scientist, Anna Palmisano, who went on to become a program manager at the Office of Naval Research and Department of Energy. We were having a conversation in the early 1990s and I was decrying the difficulty in getting funding for ‘pure’ science. Anna’s retort was that when research funds are limiting, the taxpaying public has an expectation that they will be directed towards problems of most interest to them. This of course goes against the grain of Vannevar Bush’s argument but I think hers is also a valid point.

This is very U.S.-centric. I have no knowledge whether ‘social contract’ resonates in other countries. I would love to read comments about the situation in other countries.

Leave a comment

Scientists and the public policy fray

The second part implicit in the “social contract” was that science was to be a search for “truth” outside of political power. Under the current U.S. administration that wall has been breached – Science is subject to political abuse and there are many bad actors out there willing to spread misinformation to advance their political agenda. But even before this, it was naïve to think that scientists would not get caught up in disputes over government policies – at the inception of this ‘social contract’ the theoretical physicists Robert Oppenheimer and Edward Teller were heavily involved in pushing their own agendas regarding the use and future development of nuclear weapons. This becomes problematic when scientists go beyond articulating upfront that they are advocates for a particular policy position and their science itself becomes colored by their advocacy. This leads to an erosion of public trust in science.

Certainly public mistrust is widespread for American institutions, but two issues – one acute and one multi-decadal have been particularly important for science: the responses to the COVID-19 pandemic and climate change.

With respect to COVID-19, my thoughts remain similar to a year ago.

It was necessary to make quick decisions in the absence of much ‘hard’ scientific evidence. But things like the 6-foot social distancing rule and mechanism of respiratory virus transmission could have/should have been intensively researched in the 100 years since the 1918 Influenza pandemic. In particular, the 6 foot rule just “sort of appeared” (Anthony Fauci testimony to Congress) – if it were 3 not 6 feet (for which there is very little difference for aerosols), schools could have remained open. There has been a very large social cost for school closures. I have seen very little that gives me confidence that the U.S. (or the world) is preparing better for the next (inevitable) epidemic/pandemic. In fact, that erosion of trust, fueled by the pseudo-science of our current Health and Human Services secretary, has led in the U.S. to lowered vaccination rates with consequent outbreaks of diseases like measles.

My trust in public dissemination of science also declined at the inception of the pandemic. As I was retired, I had the time to run down claims made in New York Times articles about COVID-19 by reading their basis in the scientific literature. It was far too common that the scientific manuscript made an exciting claim that was not supported by the data, and the journalist accepted this claim at face value. In my opinion, this is the synergistic consequence of scientists seeking to increase their visibility by high-impact claims and journalists seeking to increase their visibility with sensational reporting.

A broad-based problem layered on top of this is the increase in scientific fraud over the past few decades. Elisabeth Bik in particular has done heroic work in this area. Her 2016 paper found about a 10X increase in the proportion of papers with image manipulations from the late 1990’s to the mid-2010’s. Sabel et al. (2025) have used a red-flag method to intuit that about 6% of biomedical publications in 2023 were likely fake. Pressures to publish are pervasive in scientific fields and throughout the world. Paper mills have arisen particularly over the past 15 years and are a part of the problem. The traditional gatekeeper system (peer review) is under tremendous strain from the volume of manuscripts submitted; furthermore it was not designed and cannot effectively eliminate fraud.

Once I had been sensitized to the handling of new science in the media, I started looking differently at news reports regarding other policy issues such as climate change. To be upfront about my views – there is no doubt that the burning of fossil fuels has led to sharp increases in atmospheric CO2 recently at a rate of ~3 ppm per year. Simple bookkeeping of fossil fuels extracted and amounts burned demonstrate that 38 billion metric tons of CO2 were released in 2024. About 50% adds to atmospheric concentration increases and we infer ~25% becomes dissolved in ocean water, thereby acidifying this habitat.

Although from first principles it is irrefutable that elevated CO2 will lead to Earth warming, accurately projecting the consequences of warming is a difficult exercise. This is often done via ‘what if’ scenarios in Global Circulation Models. I heard the authors of these models discussing them about 15 years ago, and what struck me was how difficult it was to analyze why, when and where GCM models differed from each other, due to the long run times and inadequate granularity. That remains true in 2026 -- running the Standard CMIP6 requires about a month for 100 year projection and significant physical (cloud dynamics, ocean currents) and systemic (feedback loops) are poorly implemented. Here is a nice microbially-relevant assessment of difficulties in parameterizing such models.

Hence, I am skeptical when I see newspaper articles such as this: “Weather extremes gripping US bear climate crisis ‘fingerprint’, experts say.” I understand that scientists-as-citizens may have very strong views on rising CO2 levels as an existential crisis for the planet, but I think Science’s standing is imperiled when imprecise methodologies are pushed beyond their limits to ‘prove’ that a particular policy is necessary. Advocacy for policies is fine, as long as it is tempered by a commitment to being an Honest Broker of Policy Alternatives.

Leave a comment

Three last thoughts brought up in the NAS forum

1. That the scientific enterprise is all about the pursuit of truth is a heartwarming idea, but it doesn’t match the lived experience of junior and senior scientists in the profession. We can see it in the stressors that lead to declines in mental health from graduate students to the professoriate. That truth-seeking premise has little to do with how university faculty are hired or how they are evaluated for promotion.

2. Science as an enterprise needs to do better in recognizing where it can have negative impacts on society -- an example was the Obama administration’s pause in gain-of-function research in 2014 after the laboratory creation of an H5N1 avian virus that capable of airborne transmission between mammals. An excellent example of a proactive approach was the Asilomar Conference on Recombinant DNA in 1975 where scientists voluntarily paused their work until they could produce guidelines for safe use. The issues are often thorny as regulation may slow individual scientist’s progress but that does not absolve Science as a whole from responsibility. This statement from a 2017 report from the U.S. National Academy of Sciences summarizes the dilemma:

“There is a recognition among some leaders in the scientific community of an informal social contract wherein scientists as individuals and the international scientific community have a shared responsibility, together with other members of society, to do their utmost to assure that scientific discoveries are used solely to promote the common good. This premise is not, however, accepted by all scientific practitioners.”

3. Heather Douglas articulated her ‘taxonomy’ of research with 5 nodes that can receive public funding:

• Curiosity-based research (often referred to as “basic” or “fundamental”).

• Engaged public research involves collaboration with specific communities or stakeholder groups that have a direct interest in the research outcomes. “This is the single-most underfunded category, and also the kind that builds the most public trust,” Douglas said.

• Mission-directed research focuses on solving clearly defined problems, often tied to public priorities such as health, energy, or national security,

• Regulatory research generates evidence that is directly used to inform policy, standards, and compliance decisions

• Big-science research infrastructure

Douglas has interesting suggestions regarding Basic Research. It is virtually impossible to predict which projects will yield novel discoveries – and analysis indicates that grant panels are adverse to funding ‘risky’ research that might produce such discoveries. Hence, Douglas proposes funding screened lotteries for this category – awards would be smaller but more widely distributed, administrative burdens on principal investigators (shorter proposals) and reviewers would be reduced. This would make it more difficult for a principal investigator to engage in long continuous runs of funded curiosity-based research . I am ambivalent about her idea – but it does preserve scientific freedom for basic research while acknowledging the inherent unpredictability of major discoveries.

Anecdote #3: Norm Pace and his colleagues developed the use of 16S rDNA gene sequences to characterize ‘who’s there’ in natural microbial communities. Norm told me that he never was awarded a grant for these studies – he just used start-up funds he got whenever he was recruited to a new institution.

Final reflections

I began this post on the social contract with the thought that the burdens were light on the scientists – just pursue ‘truth.’ But I come to realize that the burdens are substantial and not easily resolved. Science (and scientists) inevitably become part of public policy debates. What is in the best interest of an individual scientist (looking for position and prestige) may not match what is best for society and hence requires strongly principled individuals. A standard comment years ago at scientific conferences was “I believe your data but not your conclusions.” But the rise of publication fakes suggests a deterioration in this moral code that we had heretofore assumed everyone followed. I admit that I am at a loss how to remedy this.

I am also at a loss regarding how to increase the U.S. public’s trust in Science. It is Kafkaesque that we have a Secretary of Health & Human Services who believes he can look at children and determine they have ‘mitochondrial challenges.’ Social media are rife with posts like this from self-proclaimed experts:

About 50% of replies mock her, but the other half clearly believe in this and that mRNA vaccines causes ‘turbo cancer.’

These bad actors will misuse whatever scientists say, but I still believe that the long game is for scientists to be scrupulous and self-critical regarding predictions, and for institutions to respond to the public’s legitimate interests at the nexus of science and policy decisions.

Today’s Moment of Zen

My wife Nancy and I ventured into Washington D.C. a week ago as the cherry blossoms were at peak bloom. Nancy got to spend 30 minutes walking around the Tidal Basin reveling in their glory, while I spent 30 minutes looking for a parking spot nearby! Finally found one about a mile away and met up with her — this is hardly unique subject matter for a photo, but I liked the gradients of light and dark in the image.

Leave a comment

I have taken a bit of a ‘sabbatical’ from this blog, but I hope to now have better regularity in my posts. In my last one on Creativity, I promised to follow up with more detailed thoughts on how biologists/ecologists explain natural phenomena via “mechanisms.” I have delved more deeply into the philosophy of biology literature on this point. What I found is that philosophers are often bemused by biologists use of ‘mechanism’ to describe their findings. To philosophers the precise meaning of words such as mechanism and whether it represents reality or a pragmatic approximation are very important. They observe a lack of clarity on both these points by biologists.

Daniel Nicholson (2013) provides a historical overview in which he claims 3 meanings of mechanism have been conflated. “Mechanism” first entered the lexicon of biology in the 17^th C with Descartes who argued that animals are automatons —complex, self-moving biological machines that could be understand from the structure and interactions of their component parts. The development of molecular biology brought new prominence to this idea. Jacques Monod described a cell as follows: : ‘this system … is … thoroughly Cartesian: the cell is indeed a machine’ (Monod, 1977, p. 108). Bruce Alberts wrote a very influential paper in 1998: The Cell as a Collection of Protein Machines: Preparing the Next Generation of Molecular Biologists. Nicholson dubs this philosophical thesis as mechanicism.

A second (metaphorical) meaning is the internal workings of a machine-like structure. The third and most recent meaning was developed by philosophers of biology over the past several decades: causal mechanism; that is, a stepwise description of how process(es) interact and operate to ‘cause’ a phenomenon. Some philosophers of science have dubbed this last meaning the ‘new mechanism’ movement and it was the subject of a book I read by Carl Craver and Lindley Darden, In Search of Mechanisms. Others (Nicholson) demur and find this label muddled as it has little/nothing to do with the thesis that cells or organisms are machine-like.

Causal Mechanisms

The distinction between the latter and the first two meanings feels very important to me. Going into this, my mental model was that intracellular processes relevant to physiological ecology (such as protein synthesis and gene regulation) were machines and cells were factories (with assembly lines of biosynthetic and catabolic pathways).
[Note that Nicholson (2019) has written critically of this idea, “Is the cell really a machine?” and suggests while the concept has been useful it also limits our capacity for deeper investigations into the reality within a cell.]

But when considering community and ecosystem ecology, a machine does not feel relevant. For the ‘new mechanist’ philosophers, a causal mechanism comprises a heuristic explanatory device of a complex system that includes a description of component objects and how they interact (in space and time) to cause an observable phenomenon.

“A mechanistic explanation consists of a ‘mechanistic model’ made up of two elements: a behavioral description (explanandum) and a mechanical description (explanans). In the case of robust mechanisms, the former is a generalization describing the mechanism’s overall repeatable behavior, while the latter shows how the parts of the mechanism and their interactions regularly bring about such behavior in virtue of their stable spatial and temporal configuration.”

A phenomenon (for example, the distribution of species’ abundance in a natural community) is to be explained by the activities of parts and their interaction with each other. These parts have spatial and temporal orientations, their activities have rates and durations. Grilli (2020) has argued that macroecological ‘laws’ describe diversity and its variation in microbial communities. I maintain that these are not laws, but rather mathematical phenomenological models of that which is to be causally explained; for example, that the mean abundance distribution of species is log normally distributed.

Vellend’s framework of 4 high-level ecological mechanisms: selection among species, drift, dispersal (immigration and emigration), and speciation might be the first level of causal explanation but note this would remain unsatisfyingly vague and would require deeper levels of causal mechanisms. Note that the mechanism is nothing like an immutable ‘law’ — the relative importance of these processes and the specific parts, activities, and interactions driving them could be very different for different microbial ecosystems. The resulting causal mechanism is modular and specific causes are difference-makers – their (experimental) manipulation affects the phenomenon.

This framework of a nested hierarchy does have utility in pursuing a research program. The highest-level begs the question as to the quantitative importance of the 4 ecological mechanisms in a specific system – measuring that would be an excellent beginning to focus subsequent research. Insights into one level are likely to provoke its decomposition into a set of lower level components and activities which can then be further investigated. The general aim is to provide ever better/more detailed answers to the question “how does it work?”

I wish that I had adopted the formalism of this approach in my research career – it is well suited to a hypothetico-deductive process and advancements in theory. It allows one to specify components that at one stage of an investigation are deemed less weighty (for example, temporal fluctuations in environmental factors) but demands return to more deeply investigate those when they address a significant gap in understanding.

An example relevant to physiological ecology. Many readers may have performed an experiment on induction of the lac operon in an undergraduate laboratory course. A graded series of inducer such as IPTG is added to an Escherichia coli culture and the activity of beta-galactosidase later measured. The results depict a saturation curve – the inference is that the lac control region acts as a rheostat in which each cell incrementally increases synthesis of lac mRNA as IPTG increases. However, when it became technically feasible to measure expression in individual cells, a completely different picture emerged. At each IPTG concentration, lac transcription remained turned off or was maximally turned on in individual cells. What changed with increasing IPTG was the fraction of cells turned on. The implication for microbial ecology is that even within one ecotype there is individual heterogeneity in how cells are responding to environmental stimuli. This certainly makes the system more complex but should also modify our thinking about how to approach experimental analyses. The only theory I can think of that has this type of heterogeneity built in is the scout hypothesis regarding dormant cells.

Causal mechanisms in ecology

As mentioned above, once one moves to community or landscape ecology of macrobes, it is difficult to imagine applying machine-like metaphors for mechanisms. Hence, elucidating causal mechanisms are most appropriate. In other words, mechanisms in ecology are not ‘things’ but rather networks of causal processes. However, the reality is that articulation of deep causal mechanisms remains rare in these fields. Macrobe ecologists have developed rich theory but often lack the appropriate data to test the theories and/or develop mechanistic explanations. Spatial scale is often a formidable problem. Two of the most elegant (and rarest) experiments in freshwater ecology were conducted at the Experimental Lakes Area in northwest Ontario. In this sparsely populated area, whole lake experiments and long-term research are tractable and seminal studies on the role of phosphorus (but not carbon) on eutrophication and quantitative assessment of the effects of acid inputs (simulating sulfur-containing acid rain) were done. One cannot imagine doing similar studies in the ocean and these experiments remains an outlier for most ecological settings.

These spatial (and temporal) impediments are a good reason to promote more research at the scales of microbial systems in a mode to identify causal mechanisms. The space, timescale for field-relevant experiments are much more tractable as is a capacity to replicate systems.

Much of macrobe (and microbe) ecology invests in discerning patterns in nature. But the end point will always be a phenomenological model (such as Grilli’s Macroecological “Laws.” They may be mathematically precise but give no insights into the causal mechanisms that generate the patterns. Developing a schema of causal processes (at multiple layers of detail) and then devising experiments and testable hypotheses is hard work and a long-term project. Microbial ecology is blessed with advanced powerful technologies. However, merely using them for increasingly sophisticated pattern recognition misses out on their real potential to get at causal mechanisms.

Leave a comment

Today’s Moment of Zen

Part of my ‘sabbatical’ entailed a three week journey through several parts of India – Odissa, Tamil Nadu and Kerala. It was my first time there and an incredible experience. Of course, one of my interests there was taking photographs and here is one I particularly liked due to the serendipity of catching a bird in flight near the goparum.

Subscribe now

I returned about a week ago from a workshop on Architectural Photography, organized by Joel Tjintjelaar, and Tobi Trumpp in Rome and Milan. I had become first aware of Joel a few years ago for his Artisan plug-in for Photoshop to aid in producing fine art black and white images. When I looked at his photo galleries and read his blog, I realized that he was thinking very deeply about this topic and felt that I could learn a lot about architectural photography from him. That was the case, but the workshop was much more than that. Joel is an intellectual who has thought about the meaning of art and how to express one’s self through that medium. He challenged us to come up with our personal definition of art. This necessitates introspection – his definition is that art is consciousness manifested.

On the plane trip back home, I thought about the relationship of his ideas on art to science. Science is not art, but both are an expression of humans about their world and the highest forms of each values creativity. I arrived home and looked into the nexus between them -- the literature on this is thinner than I had anticipated. Both require a mastery of technical methods and conceptual tools that produce familiarity with norms to creatively question them and advance the discipline.

For both arenas, the word ‘intuition’ popped up. However, the precise meaning of this for science or art felt rather murky. For the filmmaker David Lynch, intuition was an essential tool, which he defined as “knowing when something is not correct and seeing a way to make it correct.” He went on to say that it is a knowingness of feeling and thinking. Whew! I am more comfortable with “insight” than “intuition” as the latter smacks of an innate ability that we may have to varying degrees. If that were the case I should just end this blog post right now. McCrea (2010) distinguished intuition as “the ability to understand immediately without conscious reasoning” whereas insight was the “capacity to gain accurate and a deep understanding of a problem; it is often associated with movement beyond existing paradigms.” On my photography workshop, Joel provided examples and exercises for becoming more introspective and finding one’s voice as an artist. I believe there are analogous activities whereby a scientist can improve their capacity for creativity, insight, and (perhaps) intuition.

Philosophical starting point

I concur with Max Planck who in his 1933 book, Where is Science Going stated:

“…there are two theorems that form together the cardinal hinge on which the whole structure of physical science turns. These theorems are: (1) There is a real outer world which exists independently of our act of knowing and (2) The real outer world is not directly knowable.”

So we as scientists conduct experiments and analyze data to better understand nature. But particularly as biologists, we recognize that each success is not an endpoint but often reveals another, deeper level of sophistication in the mechanism. As microbial ecologists, we deal with nonlinear, dynamic, complex systems in which the ‘rules’ are often not deterministic but depend upon previous histories and future contingencies. And there is the constant assessment whether any material we experiment upon (whether a model system or sample collected from the habitat is ‘real’).

Leave a comment

Paths to deeper consciousness

My search included re-reading W. I. B. Beveridge’s 1957 book, The Art of Scientific Investigation. I remember reading it as a young scientist and was really energized by it. Fifty years on some parts are dated but it remains an articulate overview of fundamentals relevant to this topic.

On becoming familiar with norms

Beveridge advises reading the literature on your research topic – but critically not slavishly! In addition, one’s thinking can become so conditioned by the constraints of that literature that it becomes difficult to generate novel ideas. One of the greats in microbial pathogenesis, Stanley Falkow, told me that he only read the literature about once a year to avoid getting too caught up in details that might restrict his thinking.

On the other hand, reading scientific literature from a broader set of related fields may promote novel ways of looking at your research problem. An unanticipated benefit I got from teaching a microbial ecology class to graduate and advanced undergraduate students was that it forced me to look at the broad literature in microbial ecology published since I last taught the course. More often than not, I came upon insights and methods of measurement or experimentation that were useful to my own research program.

In summary, to improve your own creativity the best thing you can do is to experience very widely. Originality often arises from finding analogies between disparate objects or ideas.

Day science and Night Science

Louis Pasteur was observing motility in the bacteria responsible for butyric acid fermentation. When the cells came near the edge of the cover slip, they stopped moving, unlike many other cultures he had examined. He inferred this was due to O₂ at the interface and hypothesized that life without O₂ was possible.

August Kekulé (1890): ‘I turned my chair towards the fire and dozed. Again the atoms were gambolling … long rows …all twining and twisting in snakelike motion. But look! What was that? One of the snakes had seized hold of its own tails. … As if by a flash of lightening I awoke; and this time I also spent the rest of the night in working out the consequences of the hypothesis.’

Francois Jacob, who won the Nobel Prize in 1965 for his work on the lac operon, coined the terms day science and night science in his autobiography, The Statue Within. Day Science entails rounds of controlled experiments testing and revising one’s hypotheses – Pasteur followed up his observation on the motility of butyric acid bacteria by experimentally demonstrating fermentative organisms proliferated in the absence of O₂ and that some were killed by exposure to O₂. It is meticulous, orderly and rational.

Night science (such as Kekulé’s dream of the benzene ring) is messier – Jacob writes:“Night science wanders blindly. It hesitates, stumbles, falls back, sweats, wakes with a start. Doubting everything, it feels its way, questions itself, constantly pulls itself back together. It is a sort of workshop of the possible where what will become the building material of science is worked out. Where hypotheses take the form of vague presentiments, hazy sensations. …

“In the endless interior dialogue, among the innumerable suppositions, connections, combinations, associations that constantly pass through the mind, a beam of light sometimes rends the obscurity. Suddenly the landscape shines with blinding light, terrifying, stronger than a thousand suns. After the initial shock begins a hard struggle with habits of thought. A conflict with the universe of concepts that governs our reasoning. One is not yet authorized to say whether the new hypothesis will get beyond its initial form of rough sketch and become refined, perfected. Whether it will withstand the test of logic. Whether it will gain admission to day science.”

A less prosaic analysis identified four stages of how ideas bubbled up in other Nobel Prize winners (by Csikszentmihalyi and Sawyer, cited in McCrea, 2010). Intense reading and research came first. Next was a period of time spent away from the bench and outside distractions, on activities ranging from sleeping to playing a musical instrument. This is where the Jacob’s “beam of light” critically flashed. Stage four was more conscious work on the problem, in fleshing out and elaborating the idea, to the point that it could become ‘day science.’

Those that appear to be particularly good at this are capable of holding many sets of patterns pertaining to a domain of knowledge in their immediate memory, and particularly prize those to which they have had an emotional reaction/vivid memory.

Kekulé (1890): ‘Let us learn to dream … but let us beware of publishing our dreams before they have been put to the proof by the waking understanding.’

Discussion as a stimulus to the mind

An important element of that first stage is not just sitting in your office or at your lab bench in isolation, but interacting with colleagues! That helps identify the current norms for the scientific culture you inhabit and provides critical input on your thoughts upon the research problem. Two of the happiest times in my research career were as a postdoc in Tom Brock’s lab at the University of Wisconsin and on sabbatical in Luuc Mur’s workgroup at the University of Amsterdam. In both places, the whole lab ate lunch together. Yes, a certain fraction of time was spent on the fortunes of the Badger hockey team or the Ajax football team but a lot was spent discussing the science we were pursuing. In a number of cases this bled over from lunch to conversations with a smaller number of people that might last another hour or two. Those were not just useful – they were fun.

Chance

Louis Pasteur: “Dans les champs de l’observation, le hasard ne favorise que les esprits prepares”

Don’t underestimate the role of chance in new discoveries. But as Pasteur knew, it takes a ‘prepared mind’ to recognize something unusual and appreciate its significance. This is where ‘day science’ can hold sway for those who actively work at the lab bench or in the field, particularly when deploying novel technical or experimental procedures. But the trick is in not only recognizing but also interpreting the clue and realizing its significance. An open mind to explore unexplained observations, imagination, and that illusive property of ‘intuition’ are of great value here.

Strategy and tactics

“Work, Finish, Publish.” — Michael Faraday (1791-1867)

Beveridge considers scientific research an ‘art’ in that it does not have explicit rules. That is, as with art it should have the freedom to be anything it wants to be, to break the rules. That is certainly easier said than done in the contemporary ‘business’ of research grant funding. A general strategy is to work towards a clear objective but keep alert for and seize unexpected opportunities (for which concentrated mental effort is required to birth new ideas).

• The two essential attributes for a scientific researcher are love of science and insatiable curiosity. [tenacity in the face of failure is a close third]

• Tactical planning is better done by the operative in the field, not by an administrative committee.

• Scientific advances tend to proceed by stages – the gathering of new information leads to a pause for synthesis/interpretation/refining of hypotheses prior to further data collection.

• Courses of action at the frontier of knowledge may depend more on personal judgment based upon insights or ‘intuition’ than purely objective and rational ‘science.’

• The extremes of scientific minds are: speculative [hypotheses in advance of facts] vs systematic [accumulate data until the hypothesis is clear]

• According to Mees, scientific discovery has three different methods of working and each requires a different kind of mind: Theoretical synthesis / Observation and experiment/ Invention

• Biological discoveries are often first recognized as qualitative phenomena, but the aim should be to refine them to a quantitative basis and to discern the mechanisms that underlie them.

I had covered “mechanism” in one of my early posts. As part of my preparation for writing this post, I read In Search of Mechanisms – Discoveries across the Life Sciences by Carl Craver and Lindley Darden. There are enough interesting ideas there that I will make my reactions to their book the subject of my next post.

Bibliography

Beveridge WIB (1957) The Art of Scientific Investigation. WW Norton and Company, New York.

Csikszentmihalyi M, Sawyer K. (1995) Creative insight: The social dimension of a solitary moment. In: Sternberg RJ, Davidson JE, editors. The Nature of Insight. Cambridge, MA: The MIT Press. p. 329–363.

Lehmann, J., Gaskins, B. (2019) Learning scientific creativity from the arts. Palgrave Commun 5, 96. https://doi.org/10.1057/s41599-019-0308-8

McCrea, S. M. (2010). Intuition, insight, and the right hemisphere: Emergence of higher sociocognitive functions. Psychology Research and Behavior Management, 3, 1–39. https://doi.org/10.2147/prbm.s7935

Quotes

“With accurate experiment and observation to work upon, imagination becomes the architect of physical theory.” — John Tyndall (1820-1893)

“The really valuable factor is intuition.” — Albert Einstein (1879-1955)

“When you have at last arrived at certainty, your joy is one of the greatest that can be felt by a human soul.” – Louis Pasteur

“One word characterizes the most strenuous of the efforts for the advancement of science that I have made perseveringly during 55 years: that word is failure.” – Lord Kelvin

Even with such men in whom the hypothesis birth-rate is excessively high, it only just manages to exceed the hypothesis death-rate by a narrow margin. – WH George

“One of the first things which the beginner must grasp is that statistics needs to be taken into account when the experiment is being planned, or else the work may not be worth treating statistically”

“ It is essential to realize the impossibility of obtaining exactly similar groups. The difficulty must be met by estimating the variability …” – WIB Beveridge

“Discovery should come as an adventure rather than as the result of a logical process of thought. Sharp, prolonged thinking is necessary that we may keep on the chosen road, but it does not necessarily lead to discovery.” — Theobald Smith (1859-1934)

“Research is fundamentally a state of mind involving continual re-examination of doctrines and axioms upon which current thought … are based.” — Theobald Smith (1859-1934)

“In science the primary duty of ideas is to be useful and interesting even more than to be ‘true’.” — Wilfred Trotter (1872-1939)

“The dispassionate intellect, the open mind, the unprejudiced observer exist … only exists in a sort of intellectualist folk-lore.” — Wilfred Trotter (1872-1939)

“But the great tragedy of Science—the slaying of a beautiful hypothesis by an ugly fact—which is so constantly being enacted under the eyes of philosophers …” – TH Huxley

“It’s not that I’m so smart, it’s just that I stay with problems longer” —Albert Einstein

Your Moment of Zen

I returned a week ago from a workshop on architectural photography held in Rome and Milan, Italy. I got there a day early so visited the Vatican Museums. With respect to the topic at hand, perhaps the most appropriate thing I saw was Raphael’s fresco, the School of Athens.

Rudyard Kipling

Subscribe now

I was corresponding with Noah Fierer on ‘traits’ that are ecologically significant for an analysis done in his lab on a ‘periodic table’ for bacteria

We agreed that iron acquisition is one such trait. I became aware of this as a young pup when reading Tom Brock’s Principles of Microbial Ecology text. Brock compared for seawater the ratio between aqueous solubility and cellular content. The highest cellular ‘concentration factor’ was for iron. This was recapitulated by Moore et al. (2013).

All microbes that use an electron transport chain to generate proton motive force require Fe for their heme-containing cytochromes. This includes both chemotrophs and phototrophs. If a microbe is fixing N₂, the demand for Fe increases substantially. Nitrogenase may comprise up to 10% of total cell protein, and each functional enzyme contains 34 Fe atoms.

As noted in the Moore figure, Fe limitation may occur in some areas of the ocean – in particular ‘high nitrate – low chlorophyll’ regions such as the Southern Ocean. This has been shown in bottle experiments where Fe addition uniquely increased primary production in water samples. As increasing primary production (i.e., CO₂ fixation) has been touted as an offset to climate change linked to the burning of fossil fuels, a large-scale experiment (SOFeX) was run in 2002 whereby a few tons of acidified iron sulfate was drizzled into two 225 km² patches in the southern Ocean. The addition did indeed stimulate a phytoplankton bloom – however biomass sedimentation to deep ocean strata was variable

Another major system in which competition for Fe is important is in bacterial pathogenesis. In humans, extracellular Fe⁺³ is typically bound to proteins such as transferrin or lactoferrin and within the cytoplasm Fe is complexed with ferritin or within heme-proteins. Hence, pathogens must out-compete hosts for Fe complexes. In fact, one of the bodies responses to infection is that the liver reduces synthesis of transferrin and sequesters Fe as ferritin in the liver. Pathogens have of course evolved responses to this – laboratory experiments with cultures show that under Fe-limited conditions, the synthesis of exotoxins produced by Corynebacterium diphtheriae, Pseudomonas aeruginosa, Escherichia coli (Shiga-like toxins), Staphylococcus aureus and Vibrio vulnificus are derepressed. An inference is that these toxins cause host cell lysis, thereby making intracellular Fe accessible. A particularly attractive target for bacteria that enter the bloodstream are red blood cells, as they contain about 70% of the body’s Fe.

Iron acquisition in microbial communities (or in association with plants and animals) entails ecological interactions. Many microbes produce siderophores, low molecular weight organic compounds that form stable complexes with Fe⁺³.

To be effective for a human pathogen, the siderophore must have a greater thermodynamic stability than human transferrin to ‘win’ the competition for iron. Similarly, in aquatic environments where different species may release distinct siderophores, the stability coefficients of each molecule will determine the winners. But there is another layer to these interactions. Not only must a species produce a siderophore, it must have an active transport system to assimilate the complex. Metagenomic analyses have shown that in addition to siderophore producers with such transporters, there are ‘cheaters’ who do not synthesize siderophores but do possess the relevant transport system. Hence, theories regarding the optimal strategies for exploitation of ‘public goods’ come into play.

In ocean waters, more than 25 siderophores were detected at picomolar concentrations even at locations where Fe is not thought to be limiting (Boiteau et al., 2016). From calculation of equilibrium Fe speciation, the fraction of Fe bound to organic ligands was generally >80% (Bundy et al., 2018), suggesting that all microbes in the community must possess some mechanism to access organic Fe complexes.

There is a reasonable output of literature at the intersection of Fe, microbes, ecology and pathogenesis – about 700 publications per year. However, in skimming through it, I did not feel there has been much forward momentum over the past 10 years in understanding mechanisms. I say that with the understanding that studying Fe limitation in cultures is technically challenging because of the presence of Fe in other chemicals.

Perhaps this topic is not considered of sufficient broad interest to microbial ecology/pathogenesis? I have a few reasons why it might be. Most habitats are considered to be limited by something other than Fe. But ‘nutrient limited’ has too facets – there is the mass of nutrient in the system which determines the maximum amount of biomass and there is its concentration which may (if low enough) limit growth rate below the maximum. Bioassay experiments (adding nutrients to a sample and measuring which stimulates growth or activity) can often find that even if one element is ‘most limiting,’ others soon run out – perhaps Fe is quickly depleted to levels that limit biomass or growth rate, particularly if the maximum rate of a cellular process (such as photosynthesis or nitrogen fixation) has been up-regulated.

What do I see as fruitful areas to pursue?

Metabolomics:

• More analyses of siderophore concentrations in a variety of habitats (using LC/MS techniques). In the ocean siderophores were at reasonable levels even when the water was not thought to be Fe-limited.

• Catalogs of siderophore ‘diversity’ found in nature vs what is known from cultivars

Chemistry

If there is competition among bacteria and between bacteria and their plant or animal hosts for iron, mediated by organic ligands for Fe, then measurement of thermodynamic stability coefficients of individual siderophores as well as ‘competition experiments’ between siderophores in pure systems should generate predictions of winners and losers in microbial communities

Genomics and physiology

Predictions of siderophore synthesis and siderophore transporters from the genomes of producers and ‘cheaters.’ Paired with challenge experiments in laboratory model systems to determine if the genomic predictions yield the predicted outcome.

Leave a comment

References

Boiteau RM et al. (2016) Siderophore-based microbial adaptations to iron scarcity across the eastern Pacific Ocean, Proc. Natl. Acad. Sci. U.S.A. 113: 14237-14242. https://doi.org/10.1073/pnas.1608594113 (2016).

Moore, C., Mills, M., Arrigo, K. et al. Processes and patterns of oceanic nutrient limitation. Nature Geosci 6, 701–710 (2013). https://doi.org/10.1038/ngeo1765

Bundy RM. Et al. (2018) Distinct Siderophores Contribute to Iron Cycling in the Mesopelagic at Station ALOHA. Frontiers in Marine Science 5. DOI=10.3389/fmars.2018.00061

Your Moment of Zen

I returned from a trip to Tibet, Nepal and Bhutan about 2 weeks ago. It was extraordinarily interesting – lots on local culture (particularly Buddhist) and architecture. I have a substantial number of interesting photos that I am working through. Here is one taken on a hilltop near Thimpu, Bhutan of a bronze Buddha that is 177 feet tall.

Subscribe now

I began with the idea I would write this great piece that illuminated the interface between microbial ecology and evolution, as one where Darwin’s concepts of evolution via natural selection could be experimentally tested in practical time periods. However, I got stuck very early on – just what is the ‘unit’ that is evolving? Plant and animal ecologists have the biological species concept upon which they can frame their thinking. There is no such concept for species of Bacteria and Archaea. The 16S rDNA gene has provided wonderful insights into the phylogeny of Bacteria and Archaea but even full-length sequencing falls short of 100% accuracy in assigning species. As an ecologist, I am most interested in distributions of ecotypes. Plant and animal ecologists speak of ecotypes as different populations of a species that live in distant and ecologically distinct environments to which they have made ‘hard-wired’ (genetic) adaptations. An important question is the extent to which this holds for Bacteria and Archaea.

Two scientists who work in this area are Frederick Cohan and Kostas Konstantinidis. Fred’s research has been directed to identifying ecotypes in natural systems and elucidating the mechanisms whereby they form and coexist. He has developed theories of how bacteria may speciate. About 20 years ago, Kostas developed the use of Average Nucleotide Identity (ANI) between orthologous genes as a means to determine the relatedness between bacteria. As genomics and bioinformatics techniques have advanced, this approach has led to some interesting findings.

Bacterial species as cohesive units

Cohan argues that species are “cohesive” because there are forces that limit genetic diversification. He posits selection and homologous recombination as those forces. Selection operates via periodic sweeps – an adaptive mutation within an individual cell confers a faster growth rate/reduced loss rate upon it and its progeny such that after enough generations, the pre-existing genetic diversity within the local population is swept away. The idea that recombination is cohesive rather than divisive accrues from the observation that the probability of recombination into the genome is proportional to the homology between the recombining segments – hence primarily incorporating sequences from within the species rather than without.

Konstantinidis’ ANI analyses provide support for the existence of cohesion – he and his colleagues find a gap, with ANI of 90-95% being rare for a pair of genomes. That is, genomes from the same species have ANI>95% whereas comparisons between 2 different species yield ANI < 90%. Interestingly, when a set of genomes from the same species are interrogated, another discontinuity is found between 99.2 and 99.8%. Does this genomic microdiversity correlate with distinct ecotypes in nature?

Ecotypes

Conventional ecological thinking revolves around competition, coexistence, predation, parasitism and metabolic interactions occurring between species. But there is abundant evidence both from natural systems and from laboratory studies that intraspecific diversity exists and is ecologically meaningful. That is to say, microbial communities likely consist not only of diverse persistent species (ANI > 95%) but also another layer of microdiversity. Even if habitats are not ‘distant’ in the sense used for plant and animal ecotypes, might there be ‘distant’ habitats on a microbial scale (e.g., the pore network in an unsaturated soil) in which communities only coalesce at slow rates or ephemeral habitats (caused by the dynamics of physical mixing in aquatic habitats and/or colonizable particles)?

Cohan (2006) has proposed a “stable ecotype model” in which a population of cells shares the same ecological niche and genome sequence divergence is purged by periodic selection events. From experiments with a laboratory microcosm, he found that mutations which led to new ecotypes were just as common as those that improved the fitness of an existing ecotype. For natural communities, Cohan has favored a sequence-first approach that would presume sequence clusters correspond to different ecotypes.

The alternative is a function-first approach: identification of ecologically relevant functional differences among isolated microbes, followed by genome analysis. Examples include the work of the Chisholm lab at MIT on high-light vs low-light adapted ecotypes of Prochlorococcus and that of the Ward lab at Montana State on temperature-adapted Syncechococcus ecotypes along a thermal gradient.

Eco-Evo

Recall that four fundamental processes affect the evolution of organsisms: mutation, gene flow, natural selection and genetic drift. The dilemma here is how to identify the biological population (the fundamental unit of evolution) in prokaryotes – the analog of interbreeding plant or animal individuals. This would be a group that has reasonable gene flow among them (but keep in mind that mechanisms may differ among species, that rates of recombination vary substantially across species, and that the extant environmental conditions (cell density, physical mixing rates, spatial isolation) may impact the rates of gene flow as well as the strength of selection). This is where I got stuck in my thinking about the interface between ecology and evolution.

But I think there is a path forward and work over the past decade illustrates a path. Shotgun metagenomics has been touted as an approach for the analysis of natural communities (Simpson et al., 2025), but I think this misses the mark for reasons they cite as technical limitations – among others Metagenomically Assembled Genomes (MAGs) are composites of individual differences in the biological population such that the information on microdiversity is lost. A better alternative would be to concentrate on single-cell genomics as was done for Prochlorococcus. Kashtan et al. (2014) analyzed >1000 co-occurring cells by flow-sorting and subsequent DNA amplification. ITS ribotyping suggested there could be >>10 co-occurring subpopulations by this criterion. Larger-scale genome sequencing delineated 5 major clades. Each of these has a distinct genomic backbone, with highly conserved core genes but also a smaller set of ‘flexible’ genes. The latter varied substantially between clades. There was also fine-scale variation between core genes and flexible gene content which illustrates astounding microdiversity within this organism in the wild.

Single-cell genomics remains a small endeavor – I find only about 25 papers per year using this approach in the 2020s. It is surprising to me that no one has followed up on the Kashtan et al. (2014) findings. Technical challenges remain but are being addressed (Hosokawa and Nishikawa, 2024); it strikes me that this approach can uniquely provide the data necessary to assess genomic microdiversity in natural populations.

Analyzing systems in which the organism of interest is present in high relative abundance will help. Hence, analysis of microbial primary producers (as in the case of marine Prochlorococcus (Coleman et al., 2007) and thermophilic Synechococcus) are excellent ideas. I would add freshwater planktonic cyanobacteria: taxa such as Oscillatoria, Aphanizomenon and Anabaena are filamentous so that each “unit” will have many cells worth of DNA to isolate and amplify. These organisms can also be found in a set of lakes that differ in their distance from one another, permitting an analysis of biogeography upon their evolution. As an example, Gollnisch et al. (2024) carried out a biogeographical study on an eukaryotic phytoplankter. Among non-photosynthetic prokaryotes, highly abundant taxa such as Pelagibacter (SAR11) would be of interest.

I am optimistic that there can be further breakthroughs in this area and that it can contribute to the interface of ecology and evolution more broadly than just in microbes. There is adequate theory – Cohan’s stable ecotype model (Cohan, 2006) and its alternatives. Further developments in single cell technologies would make tests of theory in natural populations tractable. These could be complemented by laboratory experiments in microbial evolution to investigate mechanisms. Macrobe ecologists think of species as fixed assemblages of genes that determine the species’ phenotype. But the process is really a dynamic one of complex interactions between evolutionary processes and environmental factors that produce an evolutionary trajectory. A working framework is that microbial communities consist of units that can be distinguished by their genome sequence clusters and that these clusters exist both at the species level (ANI > 95%) and also as intraspecies ecotypes. Prokaryotic ecotypes may differ from those in plants or animals in that they are not necessarily geographically isolated but co-occur as a consequence of niches structured either in space or time. Experimentally pursuing the interface between ecology and evolution in prokaryotic populations has advantages due to large population sizes, multiple mechanisms of genetic exchange (even across taxa), and relatively small size of recombined sequences. These can accelerate the formation of both new clades and the pace of improved adaptation to the environment.

Analyzing systems both in the wild and under controlled laboratory conditions can identify principles regarding (i) habitats and/or taxa for which recombination rates greatly exceed mutation rates, (ii) when horizontal gene transfer is an important driving force or (iii) how environmental factors such as cell density (biofilms/GI tract vs the dilute ocean) affect the relative importance of recombination vs mutation in the organisms’ evolution and ecology.

This topic raises several very important philosophical issues:

• What is (and how to determine) the fundamental evolutionary unit (the ‘biological population’) for Bacteria and Archaea?

• Are the mechanisms of gene flow in prokaryotes analogous enough to those in eukaryotes to make the former relevant to the interface between ecology and evolution of the latter?

• What technical approaches are best suited to identify principles at the ecology/evolution interface?

• Given that the topic interfaces two silos of biology (ecology & evolution), what funding programs should be supporting research in this area?

  Leave a comment

References

Allewalt JP et al. (2006). Effect of Temperature and Light on Growth of and Photosynthesis by Synechococcus Isolates Typical of Those Predominating in the Octopus Spring Microbial Mat Community of Yellowstone National Park. Appl Environ Microbiol 72: 544-550. https://doi.org/10.1128/AEM.72.1.544-550.2006

Cohan, F.M. (2006) Towards a conceptual and operational union of bacterial systematics, ecology, and evolution. Philos. Trans. R. Soc.Lond. B Biol. Sci. 361, 1985–1996

Coleman, ML. et al. (2007) Code and context: Prochlorococcus as a model for cross-scale biology. Trends Microbiol 15: 398 - 407

Gollnisch R et al. (2024) Single-cell genomics of a bloom-forming phytoplankton species reveals population genetic structure across continents, ISME Journal 18: wrae045. https://doi.org/10.1093/ismejo/wrae045

Hosokawa, M., Nishikawa, Y. (2024) Tools for microbial single-cell genomics for obtaining uncultured microbial genomes. Biophys Rev 16, 69–77 (2024). https://doi.org/10.1007/s12551-023-01124-y

Kashtan N et al. (2014) Single-Cell Genomics Reveals Hundreds of Coexisting Subpopulations in Wild Prochlorococcus.Science344,416-420. DOI:10.1126/science.1248575

Simpson AMR et al. (2025) Investigating bacterial evolution in nature with metagenomics. Curr Opin Microbiol 87: 102654. https://doi.org/10.1016/j.mib.2025.102654.

Today’s Moment of Zen

I am leaving tomorrow for a 3 week tour of Tibet, Nepal and Bhutan. I am hoping to personally capture some Zen on this journey.

Subscribe now

The poster children for this topic are the microbes found in deep subsurface sediments that have been buried for millions of years, the Intraterrestrials.

Their ‘generation times’ have been estimated at up to a thousand years. Their existence challenges our understanding of the minimum energy flux necessary to remain alive (that is, the physiological maintenance energy). Lloyd and Steen (2025) recently commented that such organisms represent another class of extremophiles – aenophiles: “species in which all cells are obligated to grow slowly for many years or longer.”

I want to delve into maintenance energy more deeply – it is a concept that seems intuititive but the literature (on cultures) is rather fraught with different mathematical derivations and meanings. Furthermore, the calculations made upon natural populations of aenophiles would suggest maintenance energy rates <1/100^th of that measured in most cultures.

Relatively slow growth

Before that, I would like to point out that almost all microbes grow ‘slow’ in their natural environments – that is, at a rate that is well below the maximum they can achieve (µ_max). Viz. Escherichia coli can double in under an hour (µ_max ~ 0.7 h^-1) under anaerobic conditions – its in situ growth rate is estimated at 5-10% of that. Relatively slow growth is (generally) inevitable and a consequence of nutrient resource use and depletion by the microbial population. There are two facets to the concept of resource limitation in nature. One is a limit on ‘standing crop’ of biomass, determined by the resource in shortest supply. In phytoplankton, the atomic ratio of C:N:P was found to be 106:16:1. Richard Vollenweider demonstrated that it is the loading of phosphorus into freshwater lakes that usually limits phytoplankton biomass.

The second facet is that as microbes grow and deplete the ultimate limiting resource, its concentration will precipitously drop to one that can no longer sustain µ_max. This relationship was described by Jacques Monod:

Microbes’ high affinities for limiting substrate often challenge analytical chemists’ ability to measure the concentration at which µ = ½ µ_max. Not many values have been reported but in chemostat cultures they range from low micromolar to even nanomolar concentrations. In situ concentrations of individual amino acids in open ocean and lake water range from ~10-100 nM and individual saccharides at 5-50 nM -- these should depress in situ growth rates far below their maximum. Most estimates of doubling times for chemoheterotrophs in aquatic ecosystems cluster in the range of 18–50 h (Staley and Konopka 1985 ; Smith and Prairie 2004). But the case of the most abundant marine chemoheterotroph Pelagibacter ubique may be a counterexample. Its µ_max on different substrates is <0.4 d^-1 (doubling time > 40 h) and it has very high affinity transporters (K_s = 1 nM for glycine betaine).

There had been sparser data for in situ growth rates of microbes in soils, but the advent of stable isotope probing with H₂¹⁸O followed by measurement of isotope content and 16S rDNA gene copy number in labeled DNA (Koch et al., 2018) has provided a means to measure these. The bulk of taxon-specific growth rates fall between 0.01 and 0.1 d^-1 (i.e., generation times between 7 and 70 d). As it seems likely that many of these taxa will have µ_max > 2 d^-1 they are growing relatively slowly in nature.

Approaching zero growth rate

“Maintenance energy” entered the microbiological lexicon during the first phase of continuous culture work beginning in the 1960s. However the mathematics and empirical determination really measure “energy substrate not used for growth.” This might include energy dissipation processes that one would not deem ‘physiological’ maintenance. AH Stouthamer (1990) was the first to think about the reactions involved in maintenance. These include:

1. Ion leaks through the cytoplasmic membrane [e.g., proton leakage dissipating proton motive force]

2. Macromolecular housekeeping: proofreading, chaperonin refolding & resynthesizing molecules that have been damaged

3. Cell motility

4. Defense against O₂ stress.

[Obviously #3 and #4 may not operate in all organisms or under all conditions.]

The original mathematical formulations for maintenance presumed that it was a constant across conditions. However if one looks at the processes involved, this is unrealistic. Measuring the actual energy consumption rates by each process is quite difficult. However, it was an advance to think of maintenance energy in thermodynamic terms (Tijhuis et al. 1993) rather than as a debit in substrate consumption or growth rate. In this way, one can calculate a basal power requirement, which Hoehler and Jorgensen (2013) proposed as a clearer term than maintenance.

The ’slowest’ growth of cultures is generated not in chemostats but in biomass recycle fermenters / retentostats, where “near zero” growth rates have been generated. The most relevant work to the deep subsurface anoxic sediments that have extraordinarily low metabolic rates are experiments with Geobacter metallireducens (Lin et al., 2009). From their analysis of maintenance energy consumption, I calculate a power of 5 X 10^-16 watts per cell. The basal power requirements derived from retentostat cultures tend to be 10-100 fold less than those found from classical analyses of chemostat cultures. But the stunning result is that these values are ~100-fold larger than power consumption for cells in sulfate-reducing deep subsurface sediments and 1000-fold greater than in methanogenic sediments! The G. metalliruducens cells had ~6X larger cell mass than subsurface microbes – this is relevant as smaller cells will have a lower basal power requirement and can function at lower external substrate concentrations. However, the very large disparity between retentostat culture and in situ measurements suggests there may be something fundamentally different in cells existing at power rates of 10^-19- 10^-20 watts per cell.

Those calculations come from rates of terminal electron acceptor (sulfate) reduction in sediment incubations (Jaussi et al., 2023) or rates of particular organic C decreases with depth (Bradley et al., 2020) in those sediments (which of course is a proxy for time). Rates were then normalized to direct microscopic counts or qPCR of 16S rDNA genes to arrive at a power per cell value. Implicit in this approach is that all cells are metabolically active. The authors have several rationales for this and believe that few cells are dormant or quiescent, but in general these are indirect and hence more direct data would be welcome. The dataset most on point is Morono et al. (2020). When they added ¹³C-acetate or a ¹³C-¹⁵N amino acid mixture to sediments that were 95 million years old, ~90% of cells were found to have significant labeling as assayed by NanoSIMS after 21 days incubation. [Note that these were oxic sediments.]

The sense of researchers into the deep subsurface is that “such observations are consistent with a deep biosphere inhabited by cells that are not dormant or dead but that instead subsist in a viable state with extremely low metabolic rate” (Hoehler and Jorgensen, 2013).

How can they do that?

The challenges that emanate for the deep subsurface microbes and the microbial physiological ecologists who study them are many. The current thoughts are that the generation times of these microbes range from years to thousands of years. Almost all the energy they can generate must go to satisfy a basal power requirement consumed for ion gradient maintenance and macromolecular housekeeping.

Jorgensen and Marshall (2016) proposed some adaptations to minimize energy losses due to dissipation of the proton motive force. Lipids with longer fatty acid chains would be less susceptible to passive diffusion of protons across the membrane. Membrane-bound ATP synthetases driven by Na⁺ rather than H⁺ ion gradients are advantageous as the former’s transmembrane permeability as >100-fold less.

However, Kempes et al. (2017) analysis suggests that macromolecular housekeeping, particularly repairing protein damage is the largest fraction of the basal power requirement. Here is some highly speculative back-of-the-envelope math on the scale at which this might be a factor.

Subsurface cells have a volume of ~0.1 µm^-3 – if they are 50% protein, this would translate to ~300K proteins of “average” MW of 30,000 Daltons. The ‘best’ data I could find on protein degradation rates were in Escherichia coli pulse-labeled with stable isotopes (Gupta et al., 2024). They did an analysis under a carbon/energy limitation at ~15% of maximum growth rate. About 16% of the proteins had half-lives < 4 h. It is of course difficult to know how to extrapolate this to very slow growing cells (it would be a great experiment with a retentostat culture!). However, given Hoehler and Jorgensen’s estimate that sulfate-reducing bacteria in the deep subsurface process 1 sulphate ion per cell per second, resynthesis of ~50,000 proteins (@ 3 high-energy phosphate bonds per peptide bond) would require 6 months’ worth of energy derived from sulfate reduction.

Leave a comment

As long as I am talking about ~P’s … there are two other potential energy sinks in macromolecular housekeeping. Proteins may become misfolded – chaperonins can correct this but at a cost of ATP (although probably 1000X less than resynthesizing the protein). Many species grown at a low relative rate contain elevated quantities of the ‘alarmone’ ppGpp. It’s effects include a reduction in protein synthesis not only by inhibiting ribosome biogenesis but also the activity of translation elongation and initiation factors. The half-life of ppGpp is thought to be relatively short and its synthesis involves ATP and GTP -- it would be of interest to know whether this regulatory scheme operates in deep subsurface microbes as it might have a high metabolic cost.

Summing up

I really enjoyed digging into the literature on deep subsurface microbes. There are tremendous technical challenges – obtaining uncompromised sediment samples and measuring chemicals, microbes and microbial activity in excruciatingly small amounts. The cohort of scientists who have been working in this area have done an exemplary job of carrying out meticulous and thoughtful research.

But of course much remains to be resolved – what they have uncovered asks fundamental questions about the minimum requirements for life to persist and the stability of the functional macromolecules microbes contain. As several of them have noted, it would be useful to have laboratory systems that approximate the conditions in the deep subsurface. Either retentostats or small microfluidic devices would appear to offer the best options at present. There have been successes in isolating fermentative bacteria from deep aquifers associated with oil or natural gas deposits – if these bear relationship to similar taxa identified in molecular surveys of energy-limited subsurface sediments they might provide an interesting resource to investigate survival at very low energy fluxes.

References

Bradley JA et al. (2020) Widespread energy limitation to life in global subseafloor sediments.Sci. Adv.6, eaba0697. DOI:10.1126/sciadv.aba0697

Foley, M.M. et al. (2024) Growth rate as a link between microbial diversity and soil biogeochemistry. Nat Ecol Evol 8, 2018–2026. https://doi.org/10.1038/s41559-024-02520-7

Gupta, M. et al. (2024) Global protein turnover quantification in Escherichia coli reveals cytoplasmic recycling under nitrogen limitation. Nat Commun 15, 5890. https://doi.org/10.1038/s41467-024-49920-8

Hoehler, T.and Jørgensen, B. (2013) Microbial life under extreme energy limitation. Nat Rev Microbiol 11, 83–94. https://doi.org/10.1038/nrmicro2939

Jaussi M, Jørgensen BB, Kjeldsen KU, Lomstein BA, Pearce C, Seidenkantz M-S and Røy H (2023) Cell-specific rates of sulfate reduction and fermentation in the sub-seafloor biosphere. Front. Microbiol. 14:1198664. doi: 10.3389/fmicb.2023.1198664

Jørgensen BB and Marshall IP. (2016) Slow Microbial Life in the Seabed. Ann Rev Mar Sci. 8:311-32. doi: 10.1146/annurev-marine-010814-015535.

Kempes CP et al. (2017) Drivers of Bacterial Maintenance and Minimal Energy Requirements. Front. Microbiol. 8:31. doi: 10.3389/fmicb.2017.00031

Koch, BJ et al. (2018)Estimating taxon-specific population dynamics in diverse microbial communities Ecosphere 9 e02090. https://doi.org/10.1002/ecs2.2090

Lin B et al. (2009) How Geobacteraceae may dominate subsurface biodegradation: physiology of Geobacter metallireducens in slow-growth habitat-simulating retentostats. Environm Microbiol 11: 2425-2433. https://doi.org/10.1111/j.1462-2920.2009.01971.x

Lloyd, K.G., Steen, A.D. Defining ultra-slow-growing extremophilic microorganisms as aeonophiles. Nat Microbiol 10: 1555–1557 (2025). https://doi.org/10.1038/s41564-025-02048-x

Morono, Y. et al. (2020). Aerobic microbial life persists in oxic marine sediment as old as 101.5 million years. Nat Commun 11: 3626. https://doi.org/10.1038/s41467-020-17330-1

Smith EM , Prairie YT (2004) Bacterial metabolism and growth efficiency in lakes: the importance of phosphorus availability . Limnol Oceanogr 49: 137-147

Staley JT , Konopka A (1985) Measurement of insitu activities of nonphotosynthetic microorganisms in aquatic and terrestrial habitats . Annual Rev Microbiol 39 : 321 – 346

Stouthamer AH (1990) Energetics of growth at low growth rates and its relevance for the maintenance concept. In Microbial Growth Dynamics: Special Publications Of The Society For General Microbiology 28: 85-102.

Tijhuis L et al. (1993) A thermodynamically based correlation for maintenance gibbs energy requirements in aerobic and anaerobic chemotrophic growth. Biotech Bioeng 42: 509-519. https://doi.org/10.1002/bit.260420415

Your moment of Zen

Aberdeen WA bridgework

Subscribe now

It has been awhile since my first post on microbial interactions. That is a consequence of two things – I got obsessed with something else for the past few weeks and furthermore I struggled with exactly what to say regarding a path forward to understanding their quantitative importance. The first part laid out some taxonomies of interactions. Neither the classical breakdown nor my own are completely satisfying in that they might suggest all categories are of equal importance in microbial community ecology. I don’t think that likely. Therefore I will focus on just three categories: interspecies competition for resources, interference competition and the unidirectional provision of growth factors. I will also make a few comments on parasitism by viruses.

The primary take home lessons are we don’t really know how quantitatively important any of these mechanisms are because it is very difficult to measure them in authentic microbial communities.

Resource or interference competition? We see the outcome (community composition) but not the process itself playing itself out. Assays of invasion resistance by biofilm communities to introduced organisms generate experimental systems that can be evaluated but are generally less complex than an in situ community.

“Cooperation” via provision of growth factors? It is technically possible to measure concentrations of putative growth factors like vitamins (B12 is present at pM levels in seawater). However it is not the concentration of these molecules that is important – it is their fluxes. The turnover time of an extracellular metabolite could be measure by adding tracer amounts (that is, small enough to not change the concentration) of a radio-labeled form and measuring its disappearance (Hungate, 1970). However, it is difficult to obtain radioisotopes with high enough specific activity to carry out tracer experiments.

The end-members of the literature on microbial interactions are (1) generation of association networks from analysis of multiple DNA sequence-based censuses of microbial community composition and (2) mechanistic analyses of pairwise interactions of cultured microbes (in a number of cases involving mutants of the same species). Network inference has been applied for about 20 years in microbial ecology and has grown in frequency with the decrease in cost of sequencing and the development of new machine learning methods. However, in the end this approach can only identify patterns – further experimental work is necessary to determine the bases for co-existence as multiple mechanisms could result in similar outcomes. Laboratory-based model systems permit delving into the details of an organism-organism interaction, but it is not possible to extrapolate to its quantitative importance in natural communities.

Readers of previous blog posts will not be surprised that I am not a fan of hypothesis-free approaches such as the network analyses that putatively identify interactions. I am intrigued by another top-down approach (Grilli, 2020), which suggests that microbial abundance and diversity across space and time conform to certain statistical distributions:

1. Abundance fluctuations are gamma-distributed

2. The variance in the abundance of an organism is related to the mean of its abundance by a power law (Taylor’s Law).

3. The mean abundance distribution is lognormally distributed.

Grilli (2020) found that datasets from natural communities fit these patterns, and subsequently that data from 100 replicate microcosms derived from a soil community could be fit with a Stochastic Logistic Model, based upon the three rules cited above (Shoemaker et al., 2025).

These observations still beg the question as to the underlying mechanisms that produce the observed distributions. Are they primarily driven by abiotic forces so that interactions among bacteria have only weak effects (Camacho-Mateu et al., 2024) or are interactions between pairs or groups of organisms centrally involved?

Delving into causal mechanisms will require experimental approaches with relatively complex communities, but what system(s) are most likely to be tractable and relevant? Microbial mats have a number of virtues in this regard (Bolhuis et al., 2014). They contain both primary producers as well as aerobic and anaerobic decomposers and can be repeatedly sampled in time. However, accessibility to a laboratory environment to conduct manipulative experiments can be an issue as they are most prominent where fauna cannot graze them: thermophilic and hypersaline environments.

The other system that struck me as tractable is biofilms. Their compact nature makes interactions more direct and experimentally facile to carry out both replications and diverse manipulations. I found myself particularly drawn to the work of Kevin Foster and his colleagues at the University of Oxford. They focus on microbial communities associated with the mammalian gut. Their thinking brings to bear not only experimental approaches but innovative mathematical modeling, with hypotheses generated not only from ecological but also evolutionary theory.

Their publications contain a wealth of insights relevant not only to biofilms but also to many other systems. These include:

1. The availability of limiting nutrients is central to the ecology of microbial communities.

Under resource-limiting conditions, resource competition is of prime importance. For chemoheterotrophs, this generally means the diverse organic carbon and energy sources that are available. For photoautotrophs, phosphorus limitation is most common but nitrogen and iron limitations may also arise. Competition for the limiting resource entails issues raised in my blog post on physiological ecology – in particular, competitors’ specific affinity for limiting substrate(s). This is a function of the number of transporters arrayed in the membrane plus the avidity with which those proteins bind the resource. In the case of chemoheterotrophs, the regulation of simultaneous use of multiple substrates is also important.

2. Negative interactions (competition and exploitation) are much more common than positive ones like cooperation.

Mathematical modeling based upon ecological theory suggests that these interactions will generally be weak rather than strong. Furthermore, in counterpoint to a general view of microbiome scientists, cooperative interactions (i.e., those with a positive feedback loop) are destabilizing whereas competition (negative feedback loops) are stabilizing.

3. Physically-structured environments (like biofilms or soil pore spaces) provide distinct niches where nominal strong competitors can both exist.

Thus, spatial structure can result in a greater diversity (at a macroscopic scale). Priority effects (who colonizes first) become important in such spatially structured habitats as ‘immigrants’ are unable to grow as the extant community can preclude access to the limiting resource. Phenomena such as resistance to invasion by human pathogens in the GI tract are likely driven by such mechanisms (Spragge et al., 2023).

Leave a comment

Interference competition

Bacteria have evolved a broad variety of short-range and long-range weapons to disable potential competitors. Short-range systems include Type-VI secretion systems and contact-dependent inhibition. Long range (diffusible) molecules are antibiotics, bacteriocins and tailocins. These mechanisms are most likely to be effective in densely packed habitats, where cells contact each other and diffusion distances are short. Bacteriocins and tailocins are typically directed against phylogenetic close relatives (which are, of course, likely competitors for limiting resource). A recent fascinating article pointed out a nexus between interference competition and resource competition – possession of a Type-VI secretion systems can ease resource competition by liberating nutrients from a lysed neighbor cell (Stubbusch et al., 2025).

“Cooperation” via uni-directional provision of growth factors

There is abundant evidence from culturing that a substantial number of microbes cannot synthesize all of the amino acids, nucleosides, or vitamins required for growth. In addition, Ramoneda et al. (2023) did a very nice analysis of amino acid biosynthetic capability from 26,000 unique Bacterial genomes and concluded that about 20% were not able to synthesize all 20 amino acids. So, how do they get on in nature? My personal (economic) prejudice is that microbes are predatory capitalists rather than socialists (however, note that there are dissenting views – Kost et al. 2023). Has cooperative behavior evolved for provision of growth factors or do they arise from other processes? McKinlay (2023) concludes that cells are not inherently ‘leaky’ but may discharge metabolites by active or passive transporters to maintain intracellular homeostasis. However, it is important to remember that these ‘public goods’ remain available for re-acquistition by the producer as well as any other prototroph in the vicinity. Much more experimental work is required to determine whether the auxotrophies that are common in bacteria are satisfied due to pairwise cooperative relationships or by ‘normal’ physiological and ecological processes such as homeostatic efflux or cell lysis.

Viral ecology

Seawater on average contains 10 million virus particles and 1 million bacterial cells per ml. This would be a respectable multiplicity-of-infection ratio if all particles could absorb to all cells (quite unlikely). But even assuming that, the kinetics of bacteriophage adsorption would be ~100X slower than that found in typical laboratory one-step bacteriophage growth curve experiments (where initial bacterial cell densities are 100X higher). So for me it remains a paradox as to why virus numbers are so high. There are relatively few experiments that bear upon this.

I did find two recent papers that experimentally addressed the issue for cases where a specific bacterium was present at relatively high densities – Prochlorococcus and SAR11 during a phytoplankton bloom (Brüwer et al., 2024; Mruwat et al., 2023). Only about 1% of Prochlorococcus cells were infected and the authors estimated phage accounted for just a few percent of Prochlorococcus mortality. Cyanophages were 2-4X as abundant as Prochlorococcus and the authors conclude that “most encounters did not result in infection.” Given that Prochlorococcus cell densities were 10⁵ per ml, I submit that there just were not that many encounters based on first order kinetics. Brüwer et al. found SAR11 densities exceeding 10⁵ per ml during a bloom and that up to 19% of the cells were phage-infected. In 5 days, SAR11 abundance declined by 90%. These are remarkable results and I hope will be followed up, in light of the relatively low SAR11 cell densities that have yet generated rapid kinetics.

References

Bolhuis, H et al. (2014), Molecular ecology of microbial mats, FEMS Microbiology Ecology, 90: 335–350, https://doi.org/10.1111/1574-6941.12408

Brüwer, J.D. et al. (2024) Globally occurring pelagiphage infections create ribosome-deprived cells. Nat Commun 15, 3715. https://doi.org/10.1038/s41467-024-48172-w

Camacho-Mateu, J et al. (2024) Sparse species interactions reproduce abundance correlation patterns in microbial communities, Proc. Natl. Acad. Sci. U.S.A. 121 (5) e2309575121, https://doi.org/10.1073/pnas.2309575121.

Grilli, J. (2020) Macroecological laws describe variation and diversity in microbial communities. Nat Commun 11, 4743. https://doi.org/10.1038/s41467-020-18529-y

Hungate RE, et al. (1970) Formate as an Intermediate in the Bovine Rumen Fermentation. J Bacteriol 102: 389-397. https://doi.org/10.1128/jb.102.2.389-397.1970

Kost, C. et al. (2023) Metabolic exchanges are ubiquitous in natural microbial communities. Nat Microbiol 8, 2244–2252 (2023). https://doi.org/10.1038/s41564-023-01511-x

Mruwat, N. et al. (2021) A single-cell polony method reveals low levels of infected Prochlorococcus in oligotrophic waters despite high cyanophage abundances. ISME J 15, 41–54. https://doi.org/10.1038/s41396-020-00752-6

Ramoneda, J. et al. (2023) Taxonomic and environmental distribution of bacterial amino acid auxotrophies. Nat Commun 14, 7608. https://doi.org/10.1038/s41467-023-43435-4

Shoemaker WR and Grilli J (2024) Investigating macroecological patterns in coarse-grained microbial communities using the stochastic logistic model of growth. eLife 12:RP89650. https://doi.org/10.7554/eLife.89650.3

Spragge F et al. (2023) Microbiome diversity protects against pathogens by nutrient blocking.Science 382,eadj3502. DOI:10.1126/science.adj3502

Stubbusch AKM et al. (2025) Antagonism as a foraging strategy in microbial communities.Science 388,1214-1217. DOI:10.1126/science.adr8286

Your moment of Zen

I got myself immersed in learning some new approaches for black and white photography — what has been termed “fine art photography.” Not that my efforts are either fine nor art. But here is an effort derived from an image taken in Athens

Precisely what is happening in the ‘microbial community’/microbiome is a topic of intense interest to all microbial ecologists. Multiple classes of interactions are theoretically plausible, but in my opinion we don’t know quantitatively how important each class is in structuring the community. Many people have their prejudices – mine is that microbes are ruthless predatory capitalists seeking to overwhelm their competitors, but I also admit I have little empirical evidence on authentic communities to support this. I want to eventually delve into why such evidence is difficult to obtain and then think about which systems might be least refractory to analysis. But that will be for part 2 of this topic – after some reflection, I realized I should address the taxonomy of ecological interactions and its application to microbial systems.

From my earliest days as a microbial ecologist, I confronted this categorization and found it mind-numbing (From Faust and Raes (2012)):

By binning every possible combination of + and – interactions (and all do exist), I didn’t feel that one gets much of a flavor as to which interactions are major players in community ecology and which are interesting sidebars. Nor does this taxonomy aid understanding of how these mechanism operate in practice.

Readers may find this equally mind-numbing, but here is my list of biotic forces that structure community composition …

A. Interspecies competition for resources

1. Molecules/ions: Transporters and catalysts on cell surfaces

2. Production of extracellular enzymes (foraging / cheaters) — Burns et al., 2013

3. Nutrient acquisition (siderophores)

B. Metabolic interactions

1. Unidirectional provision of growth factors – amino acids and vitamins

2. Syntrophy (thermodynamics): Interspecies H₂ transfer, anaerobic methane oxidation

3. Cross-feeding

   Redox cycles (particularly in gradient ecosystems): Fe⁺³/Fe⁺² , SO₄^-2/S^-2

   Many mutualisms fit under this rubric – e.g., Rhizobium/legumes

4. Sequential utilization (metabolic feeding):

   Nitrifying bacteria

   Xenobiotic degradation: Anaerobic reductive dechlorination of PCB’s followed by their aerobic mineralization

5. Modification of local microenvironment

   Reduction in pH by lactic acid bacteria

   At an extreme, leads to succession (in which the modifying organism becomes extinct)

C. Trophic level interactions

1. ‘Microbial loop’ in ecosystems (predation by protists, rotifers and nematodes as well as predatory bacteria)

2. Parasitism (Viruses)

D. Allelopathy (chemical warfare) / Interference competition

1. Bacteriocins, tailocins, type VI secretion systems and contact-dependent inhibition (Booth et al., 2023)

2. Plant defense compounds: phytoalexins, phenolics, terpenoids, alkaloids, flavonoids, and tannins

E. Chemical communication / signaling

1. Interspecies quorum sensing

2. Plant-microbe: Root nodules

F. Structural

1. Coaggregation in biofilms

2. Microbial mats

3. Lichens / mycorrhiza

Any topics missing?

Leave a comment

Interspecies competition for nutrient resources is likely to be widespread and the most important ‘interaction’ among microbiome members, particularly to the extent that new phylotypes immigrate into the system or physicochemical conditions ‘pulse’ (Odum et al., 1995) to create variations in selective forces. Competitive outcomes when a nutrient concentration becomes growth-rate limiting entails cell membrane-bound or extracytoplasmic enzyme activities, such as the affinity of transport proteins for limiting substrates (Button, 1984), or the production of cell-surface associated or extracellular enzymes for macromolecule hydrolysis (Burns et al., 2013). Interspecific competition could also involve the production of extracellular nutrient acquisition molecules such as siderophores (Andrews et al., 2003). Particularly intriguing is the competition for soluble Fe between human transferrin and pathogens such as Pseudomonas aeruginosa, Yersinia pestis, Staphylococcus aureus and Eschericia coli.

However the catalog of microbial interactions goes far beyond simple resource competition and includes a variety of metabolic interactions. A straightforward case is where an ecotype is incapable of biosynthesis of an essential cell component (amino acid, nucleoside or vitamin). Many cultivated isolates from diverse habitats cannot grow on media with an energy source + mineral salts and bioinformatic analysis of genomes revealed ~20% were missing the genes for biosynthesis of one or more amino acids (Ramoneda et al., 2023). It remains an open question as to the source mechanism whereby these requirements are met – about 10% of aquatic primary production measured by short-term assays has been found dissolved in the aquatic milieu (although the molecular makeup of this fraction is unknown). Is ‘excretion’ by phototrophs (and chemoheterotrophs) by active (intentional) excretion, just a ‘normal’ inefficiency in retaining cytoplasmic molecules or release from cells whose cytoplasmic membrane has been damaged? Using sophisticated analystic methods, in situ B vitamin concentrations have been measured at pM levels in seawater (Sañudo-Wilhelmy et al., 2012) — but remember that the concentration depends not only upon excretion by some cells but also upon the affinity of other cells’ transporters for the molecule.

Other interactions are based on thermodynamic exigencies, as occur in syntrophic associations in which the fermentation of organic acids or alcohols is endergonic unless the metabolic end product H₂ is maintained at a low partial pressure by H₂-consuming microbes. Metabolic cross-feeding can occur in a number of ecosystems -- in gradient ecosystems such as thermally stratified lakes or microbial mats, redox cycling between oxidized and reduced forms (for example, sulfide and sulfate, or ferrous and ferric iron) can drive the interaction of either a chemoautotroph or photoautotroph that oxidizes the reduced molecule with a chemoheterotroph that uses the oxidized form as a terminal electron acceptor (Jorgensen et al., 2019). There are also cases of sequential utilization of substrates. A classic case is Nitrosomonas – Nitrobacter interaction in the oxidation of ammonia to nitrate. In addition, xenobiotic catabolism can occasionally involve multiple organisms (Bull and Slater, 1982).

In addition to metabolic interactions, secondary metabolites produced by one organism can serve as toxicants or chemical cues for microbes. Bacteriocins are one class of microbial defense systems. These small proteins are bactericidal for a narrow range of bacteria closely related to the producing strain. They have been found in all major lineages of Bacteria and some Archaea, and detailed analysis of their evolution and biochemistry have been carried out in a few organisms. Their ecological importance can be inferred from the broad abundance and diversity detected within cultivated microbes. However, the specific roles these molecules have in structuring microbial communities is unclear (Riley, 2002). Plants also can produce a broad array of secondary metabolites which have inhibitory effects not only against animal or other plant species but also members of the microbial community.

Not only do some organisms produce inhibitory chemicals, there are intriguing cases where chemicals serve as signals to an interacting partner. This is true for quorum sensing (Banerji et al. 2020) and also in plant interactions with microbes, in which they exude chemical cues that attract and stimulate specific microbial groups, as in the case of the legume-rhizobium symbiosis.

Physically-structured habitats such as biofilms and mats provide massive opportunities for interactions, including ones based on physical contact or construction of extracellular matrices. In my second post on microbial interactions, I want to dive more deeply into how biofilms could provide a great system to better quantitatively understand how interactions play a role in microbial community ecology – the dental plaque is an excellent example of this (Kuramitsu et al., 2007).

In ending, I would be remiss to not point out that properties unique to microbes yield yet another vast potential for interaction – horizontal gene transfer plus relatively rapid growth rate can produce evolutionary dynamics that far outpace any found in animal or plant communities. Microbes provide an avenue for experimental analysis of evolutionary ecology in a way extraordinary difficult for other organisms. That will be a topic for a future post.

References

Abisado, RG et al. (2018) Bacterial Quorum Sensing and Microbial Community Interactions. MBIO 9: e02331-17. DOI 10.1128/mBio.02331-17

Andrews SC et al. (2003) Bacterial iron homeostasis. FEMS Microbiol Rev 27: 215-237

Banerji, R et al. (2020) Role of interspecies bacterial communication in the virulence of pathogenic bacteria. Crit Rev Microbiol 46: 136-146

Booth SC et al. (2023) The evolution of short- and long-range weapons for bacterial competition Nature Ecology & Evolution 7: 2080–2091.

Burns RG et al. (2013) Soil enzymes in a changing environment: Current knowledge and future directions. Soil Biol Biochem 58: 216-234

Button DK (1984) The physical base of marine bacterial ecology. Microb Ecol 28: 273-85. doi: 10.1007/BF00166817.

Faust, K., Raes, J. (2012) Microbial interactions: from networks to models. Nat Rev Microbiol 10: 538–550. https://doi.org/10.1038/nrmicro2832

Jorgensen, BB et al. (2019) The Biogeochemical Sulfur Cycle of Marine Sediments. Front Microbiol 10. DOI10.3389/fmicb.2019.00849

Kuramitsu HK et al. (2007) Interspecies interactions within oral microbial communities. Microbiol Mol Biol Rev 71: 653-670. doi: 10.1128/MMBR.00024-07.

Odum WE et al. (1995) Nature’s pulsing paradigm. Estuaries 18: 547-555

Ramoneda J et al. (2023) Taxonomic and environmental distribution of bacterial amino acid auxotrophies. Nature Communications 14: 7608

Riley, MA and Wertz, JE (2002). Bacteriocins: Evolution, ecology, and application.

Ann Rev Microbiol. 56: 117-137

Sañudo-Wilhelmy SA, et al. (2012) Multiple B-vitamin depletion in large areas of the coastal ocean. Proc Natl Acad Sci USA. 2012;109:14041–14045. doi: 10.1073/pnas.1208755109.

Slater JH and Bull AT. (1982) Environmental microbiology: biodegradation. Phil Trans Roy Soc B: 297: 575-597. https://doi.org/10.1098/rstb.1982.0063

Your moment of Zen

I’ve been spending some time learning new ways of processing black and white images — here is a current product that is also colorized, taken at the Romanesque Abbazia Sant’Antimo near Montalcino in the Val d’Orcia, south of Siena, Italy

Subscribe now

A break from normal TLaM service today …

As a retired scientist in the United States, I have watched what has happened to scientists and their institutions under the Trump Administration these past three months. My former colleague and continuing friend David Asai has written an essay on how scientists and educators should resist the forces unleashed by Trump and his acolytes to make sure that the enterprise of science is available to everyone who makes up this country.

In the event there are readers who do not reside in the United States, here is a short summary of actions taken or planned by the Trump Administration.

1. The removal of “DEI” (Diversity, Equity & Inclusion) staff from government offices and material from government websites. They also have threatened (and in some cases initiated) financial and legal actions against universities unless they remove “DEI” from their institutions.

2. The removal of webpages on and funding for research on climate within the federal government and to research universities.

3. Under the cover of “investigating anti-Semitism” or eliminating “racial preferences” (against white men) on campuses, Trump has withheld funding (largely in support of scientific research and totaling hundreds of millions to billions of dollars) to universities such as Columbia, Harvard and others unless they agree to a broad list of demands that eviscerate a university’s constitutional rights and any notion of academic freedom. Columbia bowed to those demands, but is still negotiating for restoration of the funds. Harvard’s President Garber published a pointed letter refusing to cave in to the government.

4. More than 1000 foreign students have had their visas revoked, thereby exposing them to immediate deportation by Immigration and Customs Enforcement officers (and at a number of institutions with the assistance of university police). In general, the revocations come from action in demonstrations or social media posts in support of Palestinians. There has been a large public outcry and the Trump administration announced it would reinstate the canceled visas.

5. Contractually funded science research grants from the National Institutes of Health, the National Science Foundation, the Department of Energy and the Department of Agriculture to many universities have been ‘suspended’, frozen, or terminated.

NSF alone has terminated ~750 grants that had awarded $739 million. On April 30, NSF staff were told to “stop awarding all funding actions until further notice.” Ongoing careers of young scientists is jeopardized and work in progress may be irrecoverable.

ADDENDUM: Grant Watch tracks grant terminations at the NIH and the National Science Foundation. The author, Scott Delaney, is to give academic researchers and their advocates a tool to resist policy decisions.

6. Proposed budgets for future fiscal years suggest the following:

• a reduction from $45 billion to $27 billion in funding for the National Institues of Health. This would entail a reduction of about 20% of scientists working within the Institutes and of course a large reduction in extramural research grants to universities and medical centers.

• the White House proposes a 55% cut to the National Science Foundation’s $9 billion budget. It is the government agency most devoted to support of basic research.
  A recent macroeconomic analysis suggests 25% cuts in research funding would reduce American economic activity to the same extent as the Great Recession of 2007-2009

• The National Oceanic and Atmospheric Administration is expected to lay off 20% of its workforce with likely effects on its weather forecasting work ahead of hurricane season. Furthermore, NOAA’s scientific research division may be abolished, with funding slashed by 27%, eliminating “functions of the Department that are misaligned with the President’s agenda and the expressed will of the American people” (AKA climate change). The primary locus of climate science, the Office of Oceanic and Atmospheric Research would be cut by 74% meaning the shutdown of climate data collection and modeling.

• Scientific research divisions within the US Geological Survey, US Forest Service and US Department of Agriculture are also slated for large budget cuts or total elimination.

  Leave a comment

‘DEI’ in science

The phrase ‘DEI’ is used as a cudgel by Trumpists as if this is a novel recent invention by those they demonize. But as Rob Bradford pointed out to me, it has been around for more than 150 years – the Emancipation Proclamation was DEI, Women’s Suffrage was DEI, school desegregation and elimination of Jim Crow laws in the South in the 1950s and 1960s was DEI. But there remains more work to do to ensure a ‘level playing field’ for all in American society. Whereas Trump and his supporters maintain that women, Black and Hispanic people with positions in society are evidence of discrimination against white males, I think there is stronger evidence that the continued under-representation of people with these visual characteristics is due to the continued presence of structural factors in society that work against their advancement based upon their merits.

The raw data suggests that this is true in science, and the point is particularly clear in biology. Biology attracts more women to graduate school and academic careers than do other science and engineering disciplines. About 50% of graduate students and 40% of postdocs in biology are women. However, only 36% of tenure-track assistant professors and 18% of full professors are women. If one looks to the ethnicity of graduate trainees and faculty at universities, there are large disparities in the proportions of Black or Hispanic Americans compared to their representation in the U.S. population.

Certainly it is within the power (if not the wisdom) of the Executive branches at the federal or state level to expunge references to ‘DEI’ from its websites. In a number of individual states controlled by Republicans, legislation has been passed and signed by governor to force public universities to fire ‘DEI’ staff and close centers devoted to the support of Black or Hispanic students. Others have ‘voluntarily’ closed such centers even without legislation. The Chronicle of Higher Education has documented 37 states where changes have occurred.

But the effects of these actions go beyond public institutions and have generated ‘anticipatory obedience.’ I mentioned my friend David Asai at the top of this post. David finished his career at the Howard Hughes Medical Institute (HHMI), as Senior Director of Inclusive Learning where he oversaw programs that provided financial support to universities developing novel ways to educate and encourage individuals from under-represented groups to pursue science at a high level. His group administered several programs, including SEA-PHAGES, created by Graham Hatfull, in which beginning undergraduates isolate and characterize bacteriophages. SEA-PHAGES now has more than 130 active member schools where 5,500+ students and 600+ faculty annually engage in discovery-based learning. Another program developed by Asai's group was Inclusive Excellence, which challenged colleges and universities to create a more inclusive science learning environment for students from all backgrounds. In early February, HHMI announced that it was terminating the Inclusive Excellence grants to 134 schools ($60M) to those institutions immediately and any mention of this Inclusive Excellence program (past or present) was scrubbed from its website (this last action is literally Stalin-esque).

HHMI gave no reason for the termination but the timing with respect to Trump’s assumption of government power makes it obvious. Note that HHMI is a non-governmental organization that gets no support from the federal government and controls nearly $26 billion in assets which it can use to support biomedical science research and education as it sees fit. My own American Society for Microbiology scrubbed from its website links to Hispanic, Black and Native American web pages in early February. There were angry letters from members and the websites were eventually restored. However, by that time the damage was done to young microbiologists who needed a strong ally that had their back in these tumultuous times but found that ally was AWOL.

David Asai wrote in his essay that “the goal of diversity, equity, and inclusion (DEI) is for individuals with different lived experiences and divergent perspectives to genuinely feel that they belong and can contribute to the success of the larger community.” I concur. As is my wont, I try to frame progressive causes also in terms of their benefits to society – excellence in science is a rare enough commodity that we cannot afford to not welcome and encourage people from a broad variety of backgrounds to pursue it to the limit of their abilities.

Resistance

Who should resist these attacks on universities and science and work to improve science culture? Well, of course everybody but here I will focus on university faculty, laboratory group leaders and department chairs -- positions of privilege which thereby demand responsibility

Universities have multiple layers of bureaucracy, but it is the faculty member who is responsible for fulfilling the fundamental missions of a university: teaching and research. Therefore, it is they who must be at the forefront of effecting change in the cultures of both the classroom and the research laboratory. Good intentions are not enough. This takes work– which involves significant time in a profession where there never seems to be enough of it. The first element of this ‘work’ is to truly listen to students and postdocs, particularly if they are coming from a world of experiences very different from your own. Secondly, be very sensitive to messages that you or others in the laboratory or classroom may be sending (even inadvertently) that someone is not a valued contributor to science.

Institutional leaders can promote this dialogue also. David Asai relates experiences he and I had at Purdue where the School of Science organized weekend retreats for faculty to sensitize themselves to classroom climate for women and PEERs*. We faculty grumbled but after the fact I viewed this as a very positive experience and stimulated me to do better at listening to those with different experiences. The Culture Box is one example of how to raise the awareness of mentors, but I reiterate that this needs to be a ‘life-long learning process’ for faculty.

* David prefers PEERs [Persons Excluded from science because of their Ethnicity or Race] to the term Underrepresented Minority. PEERs are 32% of the U.S. population, 37% of undergraduates, 21% of STEM bachelor’s degrees, 12% of STEM PhDs, and 6% of tenured STEM faculty

On being an ally

Universities, organizations and their constituent departments and staff have proclaimed themselves allies with statements in support of Black Lives Matter and acknowledgment of the occupation of Indigenous Lands. If left at that, these become performative and meaningless. Perhaps an institution goes farther and appoints a Vice-President of Diversity, Equity and Inclusion. But we have seen recently that as soon there is a headwind, these allies may shrink away.

Faculty, postdocs and students can serve as staunch allies at the local level in their research laboratories by listening to and supporting PEERs. Vocally support institutional leaders who stand up to assaults on academic freedoms and resist exclusionary policies, such as President Garber at Harvard. Question leaders who appear to give in to those assaults and ask if it is really required by legislation (not executive orders) or is ‘anticipatory obedience.’ Support efforts for collective resolve by universities, so that individual institutions cannot be so easily targeted. This can require real moral courage by individuals and institutions and comes from a deep understanding of what the most cherished values demand.

Awareness / listening /interventions are important not only for PEERs. Mental health issues in research laboratories among students, postdocs and faculty colleagues are widespread. Fifty percent of 4,000 survey respondents reported struggling with depression and anxiety and >65% had witnessed bullying or harassment in the laboratory . Laboratory leaders are very unlikely to hear about these things unless they are completely open to listening and aware of symptoms so that they can encourage those suffering to seek help. Faculty who serve on graduate examining committees are in the best position to detect cases where there is a problematic relationship between mentor and student and initiate remedial action. The sociologist Erin Hatton’s book Coerced: Work under Threat of Punishment analyzed 4 cases where the power relationship between supervisor and ‘laborers’ can lead to toxicity: prisoners, welfare recipients, college athletes and … graduate students.

Reforming how introductory biology is taught

The status quo has not been working, not only for PEERs but also for many white males and females. More than 50% of first year students in STEM majors switch before their senior year. This despite large investments in ‘student success’ initiatives, academic counseling, enrichment activities and programs aiming to increase access and success of PEERs in biology – none have had measurable effects. Resistance to change often comes from the faculty themselves – deciding upon and implementing a new approach would take significant time and effort. I get it – but biology has experienced revolutionary changes this century in technology, methodology and its interactions with chemistry, physics, math and environmental sciences. How recently was the curriculum at your institution been revised to reflect science as a modern process, rather than as history? Does it remain a silo where students are expected to figure out on their own how to integrate its knowledge with that from other natural sciences and mathematics?

This cannot be mediated at the individual level but requires department and multi-department interactions. I am fully aware that no matter how difficult some of the challenges above are, this is probably the most refractory to change. Certainly individual champions can be catalytic but require not only pedagogical but also political skills to navigate the academic minefield.

Formal laboratory courses provide a rich opportunity to excite students about the power of discovery in science. This requires innovations such that classroom periods are not cookbook exercises but research experiences that emphasize curiosity and critical thinking. May this be more ‘expensive’ in terms of supplies and staff effort? Sure – but what value is lost by the current high dropout rate of biology majors?

Putting on my ecologist hat for a moment, university field stations also have great potential to engage students in formal and informal research experiences. Residence at the station via formal courses or research work will inevitably expose a biology student to chemists, engineers, hydrologists, geologists, soil scientists who are also carrying out work in environmental sciences. I spent quite a bit of time early in my career doing research at several field stations and I think that is what really turned me on to doing interdisciplinary research. These stations are great resources for their institutions – unfortunately, they are at the end of the ‘pipeline’ for financial support. Without a strong on-campus champion, facilities may erode due to benign neglect.

Leave a comment

Essays by David Asai

Asai, D.J. 2025. Inclusive Science Education Is Not Zero-Sum. Issues in Science and Technology 41: Spring issue.

Asai, D.J. The Little Red Hen and Culture Change. Plant Science Bulletin 67(3): 174-180.

Asai, D.J. 2020. Race Matters. Cell 181: 754-757.

Asai, D.J. 2020. Excluded. Micro. Biol. Educ. 21(2): 1-2.

Asai, D.J. To Learn Inclusion Skills, Make it Personal. Nature 565: 537.

Asai, D.J. The Little Red Hen and Culture Change. Plant Science Bulletin 67: 174-180.

Asai, D.J. 2020. Race Matters. Cell 181: 754-757.

Asai, D.J. 2020. Excluded. Micro. Biol. Educ. 21(2): 1-2.

Poodry, C.A. and D.J. Asai. 2019. Questioning Assumptions. CBE-Life Sci. Educ. 17:es7.

I’ll be back to microbial ecology soon — thinking about the extent to which we know microbes in communities are socialists or predatory capitalists. But for now …

Your Moment of Zen

Along the coast of Chile, there is a temperate rain forest (1 - 6 m of precipitation per year) that I visited in March. The setting (geographically and geologically) is similar to the Hoh Rain Forest in Washington State. But the plants in this Valdivian Rain Forest are different …

Subscribe now

Subscribe now

A little history. As a high-school student I loved mathematics. I think this was due in large part to the enthusiasm and rigor of my teacher. However, a couple of semesters at university where I was taught by graduate teaching assistants wrung all of that love out of me and I did not progress beyond calculus. I regret that, as I later saw how facility in mathematics could have improved my science. About 30 years ago, I read the esteemed limnoecologist WT Edmondson’s reflections “The Uses of Ecology: Lake Washington and Beyond.” As a high school student working in GE Hutchinson’s lab at Yale prior to entering college, he asked Hutchinson for advice on courses. Hutchinson’s answer was “keep taking math courses until you fail one.” Sage advice – mathematics is a tool that you should develop to the limit of your abilities. (Although I imagine the pre-med students I advised at Purdue would not have been swayed by this argument).

This post directly follows my thoughts on theorizing. A theory need not be mathematical, but if it is expressed in mathematical language, it has the virtue of generating quantitative predictions which can then be empirically tested (that is, Popper’s method of advancing science). Some mathematical models may be solved analytically, but more popular are simulation models in which a system of ordinary differential equations are solved via numerical integration.

History

The ’classic’ models of the early days of ecology were the Lotka-Volterra predator-prey equations and the Verhulst (logistic) growth equation, both developed in the 1920s. There is a generalized form of Lotka-Volterra for any number of species that compete or otherwise interact with each other. This comprises a set of ordinary differential equations (one for each species) in which the maximum growth rate of the species is provided and there is an interaction matrix that describes the interactions between each pair of species. Whereas the predator-prey equation generates limit cycles, the more complex interaction matrices of the generalized equations may result in limit cycles, point attractors, or chaos.

Phase plane of Predator vs Prey in a Lotka Volterra model, illustrating limit cycles whose amplitude depend upon initial conditions of prey and predators (X0).

The logistic growth equation remains broadly used in ecology, although in my opinion the Monod equation

dX/dt = X *µ_max * S/(K+S)

coupled with a specific accounting of limiting substrate consumption

dS/dt = -(µX/Y)

is far superior.

[X=biomass, S=substrate concentration, µ=instantaneous growth rate, K=concentration of S necessary for 0.5 µ_max and Y=yield biomass per unit substrate]

Leave a comment

The latter half of the 20^th Century produced some important advances in theoretical ecology and modeling. Howard T. Odum was a driving force behind the development of “systems ecology” and simulation models of the energy flows through ecosystems. These began as electrical circuit diagrams (he used batteries, wires, resistors and capacitors as analogs of ecosystems in his earliest work!) but developed into a series of linked ‘reservoirs’ with regulated inflows and outflows. His work exemplified the top-down approach to ecology and his 1994 book, Ecological and General Systems, remains a useful compendium of the mathematical, kinetic, and energetic bases for the systems ecology approach.

In my opinion, Robert May had an outsized influence on mathematical approaches to ecology. Trained as a theoretical physicist, the work of Hutchinson and Macarthur led him to apply his mathematical expertise to fundamental questions in ecology starting in the late 1960s. His 1973 monograph, Stability and Complexity in Model Ecosystems, explored the diversity-stability debate and demonstrated that stability was far from a given in randomly-assembled communities. This opened an entire field regarding what characteristics make systems robust and stable and what types of conditions will produce transitions in community composition. The conventional wisdom at the time was that shifts in community composition were driven by stochastic changes in exogenous factors. However, his analysis of simple sets of difference equations showed that ‘internal’ dynamics could of themselves generate highly complex community responses, venturing as far as what is now called chaos. The generality of this idea remains unsettled, although nonlinear dynamics continues to be an important basis for modeling the complex behaviors observed in natural communities. He and Sugihara (1990) later developed an approach to make predictions regarding the trajectories of nonlinear dynamic systems from time series data. Sugihara call this approach Empirical Dynamic Modeling.

Models in Microbial Ecology

The record is rather thin here. Of course, microbial ‘compartments’ are part of systems-level ecosystem models, although their representation is at a very coarse level. Biofilms are a natural for Individual-Based Modeling, given their importance in dental caries and a variety of engineered ecosystems. In this approach, each individual cell is modeled in space and assigned ‘rules’ that govern its behavior. This is intriguing as it can get to the microenvironment scale. However, these models can quickly become complex and hence computationally expensive unless the rules (and spatial scale) are quite simple. Jan-Ulrich Kreft and his colleagues have developed a simulator iDynoMics (Cockx et al., 2024) that can be downloaded from GitHUB.

Genome-scale metabolic flux models came into vogue as complete genome sequences of bacteria became readily available. These are based on flux balance analysis (FBA), developed by the chemical engineer Bernard Palsson and his colleagues. From the genome, a matrix of reactions is constructed. Constraints are imposed and a solution is obtained to maximize an objective function (for example, growth). The output is the flux of metabolites through the various metabolic pathways. This is a great approach for questions in chemical engineering reactors, where microbes may be in steady-state or balanced growth conditions. However, note that the FBA only produces a solution at steady state – that almost never occurs in natural microbial communities.

Leave a comment

In my opinion, these metabolic flux models miss the mark by focusing on containing all the details (enzymes) of metabolism, many of which are common to all organisms. Going back to the mantra of Clements that ecology is the study of stimulus and response, I believe that the focus should be on the regulation of cell activities by the extant physico-chemical conditions. Based on the collaborations I had with Doraiswami Ramkrishna (Ramki), an eminent chemical engineer at Purdue, on modeling a microbe in a bioreactor, I came to believe that the key was to simulate the effects of regulation while minimizing the details. His approach is cybernetic modeling which implicitly represents regulatory mechanisms, and calculates their impact via cybernetic variables that constrain metabolic processes to achieve an (assumed) goal (for example, maximize growth rate or minimize death rate). Of particular importance in environments containing limiting or transiently-available concentrations of heterogeneous nutrient resources would be simulating regulation of different transporters. Ramki’s cybernetic models contained ‘enzymes’ but these were lumped entities (for example, protein synthesis capacity) to reduce model complexity.

Hyun-Seob Song led the generation of a review article that covered all of the frameworks used to model microbial communities. Although 10 years old, it remains a good overview of the discrete approaches.

Where are we?

I find myself disappointed that mathematical modelling of microbial populations or communities has not advanced further. I found only 65 publications in what I consider high-quality journals over the past 5 years. Some of the reasons go back to my science-of-microbial-ecology-science post: current funding mechanisms do not particularly value this type of research. I also view the genome-scale flux-balance approach an intellectual dead end for microbial ecology – but almost half of those 65 publications used this.

Progress requires effective intellectual engagement between microbial ecologists and mathematicians. It would be rare for a single individual to function at the highest level in both ecology and mathematics. I go back to my collaborations with Ramki and his students at Purdue. Yes, the topic was a simpler one – modeling the activity and growth of a microbe in a bioreactor. We would sit and discuss the principles of microbial physiology relevant to the model and the mathematics behind implementing it. The students were very receptive to learning more about microbial physiology and I got to the point where I could read a page of ordinary differential equations and glean what they expressed. This required serious work on both sides to understand where they were coming from.

I had mentioned Hyun-Seob Song’s review of modeling of microbial communities earlier. He is a former colleague who exemplifies the type of traits necessary for successful collaborations in this area. He has a deep understanding of mathematics and a willingness to work together with microbiologists, biogeochemists and environmental scientists to advance this area. He is now a faculty member at the University of Nebraska-Lincoln, and I find the directions he is pursuing with colleagues of great interest.

A premise I developed in earlier posts was that microbial ecosystems are complex systems™️, in which dynamics are nonlinear and driven both by external perturbations (regular or stochastic) and internal (biological) dynamics. If that is so, it feels essential to make the mathematics of nonlinear dynamics and concepts such as limit cycles, attractors, bifurcations and regime shifts an essential part of modeling efforts — see Fujita et al. (2023) for a recent example.

References

Cockx BJR et al. (2024) Is it selfish to be filamentous in biofilms? Individual-based modeling links microbial growth strategies with morphology using the new and modular iDynoMiCS 2.0. PLoS Comput Biol 20(2): e1011303. https://doi.org/10.1371/journal.pcbi.1011303

Fujita H et al. (2023) Alternative stable states, nonlinear behavior,and predictability of microbiome dynamics. Microbiome 11:63. https://doi.org/10.1186/s40168-023-01474-5

May, RM (2001, revised) Stability and Complexity in Model Ecosystems. 301 pp. Princeton University Press;

Odum, HT. (1994) Ecological and general systems : an introduction to systems ecology. 644 pages. University Press of Colorado, Niwot, Colo., ©1994

Song H-S et al. (2014) Mathematical Modeling of Microbial Community Dynamics: A Methodological Review. Processes 2:711-752. doi:10.3390/pr2040711

Sugihara, G., & May, R. M. (1990). Nonlinear forecasting as a way of distinguishing chaos from measurement error in time series. Nature, 344: 734–741. https://doi.org/10.1038/344734a0

Some general overviews on modeling in microbial ecology

Kreft JU et al. (2017) From Genes to Ecosystems in Microbiology: Modeling Approaches and the Importance of Individuality. Front. Microbiol. 8:2299. doi: 10.3389/fmicb.2017.02299

Martinez-Rabert E et al. 2023 Multiscale models driving hypothesis and theory-based research in microbial ecology. Interface Focus 13: 20230008. https://doi.org/10.1098/rsfs.2023.0008

Vallina SM et al. (2019) Models in Microbial Ecology. Encyclopedia of Microbiology 4^th edition, pp.211-246. Elsevier Inc.

van den Berg, N.I. et al. Ecological modelling approaches for predicting emergent properties in microbial communities. Nat Ecol Evol 6, 855–865 (2022). https://doi.org/10.1038/s41559-022-01746-7

Widder, S. et al. (2016) Challenges in microbial ecology: building predictive understanding of community function and dynamics. ISME J 10, 2557–2568. https://doi.org/10.1038/ismej.2016.45

Your moment of Zen

My procrastination in producing this post was a combination of being away for 3 weeks in Patagonia and recovering from that expedition. Patagonia is a fascinating area – as someone who lived in and visited low population density areas of Western North America, it was fascinating to travel ~4000 km by ship and see very few towns. Well, there were some occasional hot spots with creatures new to me — in this instance Emperor Cormorants

Subscribe now

My inspiration for this blog has been the eminent microbial ecologist Jim Prosser. Jim has written a series of thoughtful pieces on the philosophy of microbial ecology over the past 17 years. That first article was entitled “The role of ecological theory in microbial ecology” in which he and his coauthors said that the underdevelopment of theory had stunted progress in microbial ecology and that an important factor was “cultural, in that the tools and disciplines of ecological theory are not part of the contemporary mindset in microbiology.” That article has certainly been widely read --it has been cited more than 700 times. However, as I reflect on the current state of microbial ecology research I cannot unequivocally assert that this cultural barrier has been overcome.

This barrier contrasts with macrobe ecology, which has a long and deep history of theory development (as well as continuing disagreements between empiricists and theoreticians on the role of each other’s approaches). My current opinion is that microbial ecology need not develop theories unique to itself but can build upon and enrich macrobe ecology theory. Theory can provide a framework to develop testable hypotheses. A weakness in macrobe ecology theories is the difficulty in testing them at the appropriate time and space scales and with adequate replication. I think there is enough congruence between the ecology of microbes and macrobes that the advantages of microbial systems can be exploited to rigorously test these theories.

However, the application of ecological and evolutionary theory to microbiome community dynamics is complicated by features of the microbiome, including horizontal gene transfer, rapid evolution and the production of public goods and antimicrobial compounds

The hierarchy of theory

I read a book edited by Scheiner and Willig (SW), The Theory of Ecology, published in 2011 that was very illuminating regarding a philosophical basis for theorizing. SW define theories as “hierarchical frameworks that connect broad general principles to highly specific models.” They propose three (loose) tiers: general theory, constitutive theories and models. In their view, general theory sets a broad framework of principles upon which constitutive theories are developed. These generate propositions (relationship or causation) to explain specific phenomena. Models are the level where Popper’s exhortation to generate ‘risky hypotheses’ and test them is fundamental. The tests may entail field experiments but may also be statistical or mathematical models. A good example of the latter was Robert May’s seminal work on the diversity vs stability debate (May, 1973). He carried out linear stability analysis of randomly constructed communities with random interaction strengths and found that diversity destabilizes community dynamics when perturbed. This was puzzling to ecologists and at odds with some empirical results. However, the controversy stimulated efforts to understand how authentic communities may not be random and what kinds of interaction structures promote persistence.

General theory of ecology

SW have put forward their proposal for a general theory that consists of 8 principles. I reproduce them here.

Eight fundamental principles of the general theory of ecology

1. Organisms are distributed in space and time in a heterogeneous manner.

2. Organisms interact with their abiotic and biotic environments.

3. Variation in the characteristics of organisms results in heterogeneity of ecological patterns and processes.

4. The distributions of organisms and their interactions depend on contingencies.

5. Environmental conditions as perceived by organisms are heterogeneous in space and time.

6. Resources as perceived by organisms are finite and heterogeneous in space and time.

7. Birth rates and death rates are a consequence of interactions with the abiotic and biotic environment.

8. The ecological properties of species are the result of evolution.

SW anticipate your immediate assessment: “The reaction of many to confirmed generalizations is, “Well, isn’t that obvious?” In reality, the answer is no. Often such generalizations are obvious only after their explication.”

On reflection, I see that the principles I set out in Big questions in (microbial) ecology for the most part fit within this framework. If there is something missing from a microbial perspective, it might be the fundamental importance of thermodynamics.

Leave a comment

Constitutent theories

Superficially, ecology looks to be theory-rich. Marquet et al. (2014) identified 78 theories in the ecological literature, although some might better fit as ‘models’ than constituent theories. Furthermore, a few are rather specific for plants or animals (e.g., life history theory) that they are not very meaningful for microbial ecology.

I did a Web of Science survey of microbial ecology papers published since 2000 (about 78,000) – only 2,000 contained “theory” as a Topic. Those most commonly mentioned were related to two of Vellend’s ecological processes that determine community composition: selection among species and dispersal. These papers rarely tested Vellend’s theory but rather used the processes as a context for their molecular censuses of microbial community composition. The metabolic theory of ecology and game theory were the next most commonly cited.

There certainly are a lot of ecological theories that could be tested using microbes– either natural or model systems. Some of the more obvious ones (based on Marquet et al.’s list) include:

Drivers of community composition:

1. Biological Invasion Theory

2. Community Assembly / Species Diversity Theory

3. Competition / Coexistence / Competitive Exclusion / Niche Theory

4. (Intermediate) Disturbance Theory

5. Insular Biogeography Theory

6. Landscape / Metacommunity / Macroecological Theories

7. Neutral Theory of Biodiversity

“Life history” strategies

1. C-S-R Triangle / Metabolic Theory

2. Dynamic Energy Budget Theory

3. Ecological Stoichiometry / Resource Ratio Theory

4. Optimal Foraging Theory

5. Trait-based Theory

Ecosystem / larger scales

1. Ecological Network Theory

2. Information Theory (see RE Ulanowicz)

3. Diversity-Stability Theory

4. Succession Theory

Another very important area to which I will devote a future blog post is Evolutionary ecology theory. This area became discernible in macrobe ecology during the 1960s to investigate the environmental factors that drive species adaptation. However, many questions remain outstanding. I think microbes can play a special role here, given the strong selective pressures that can be applied at a small spatial scale, their relatively rapid rate of growth which can fix advantageous mutations and their added capacity for horizontal gene transfer.

Models

The third tier of SW’s hierarchy are Models. They will be the most specific of the tiers and may be pertinent to a particular mechanism or system. They may very well be conceptual or abstract, and comprise statistical, analytical or computational approaches. The intent is to generate predictions that can be empirically tested. These models are not a description of reality but rather our assumptions about reality. Hence, it is appropriate that the model evolves as it is empirically tested.

Per some of Jim Prosser’s comments over the years, the effort required to take your study-of-interest and tune its experimental details to test a hypothesis/model that falls underneath a constitutive model is not enormous and would markedly enhance the impact of your work.

An example: Diversity-stability theory.

The question whether biological diversity confers stability has been a contentious one for decades. A significant part of the problem are the diverse ways in which ‘stability’ has been defined and measured (Donohue et al., 2016). Here I take as my conceptual model that higher microbial diversity (and in particular functional redundancy) will result in greater stability of ecosystem function (measured as key process rates such as primary production or degradation of organic matter), observed as resilience to or recovery from environmental perturbations.

The tests and limits of this model could entail natural systems in which there are environmental gradients of physicochemical factors such as temperature, pH or salinity – hot spring mat communities provide a good example for temperature. Alternatively, natural communities can be diluted to reduce diversity (this will tend to selectively exclude ecotypes found in low relative abundance) and then reinoculated into a sterilized natural sample or a synthetic analog of the habitat. Lastly, synthetic communities of differing diversities could be constructed from cultures sourced from the habitat of interest.

In my ‘dream’ world, there is the capacity to continuously monitor system function (for example, an oxygen microelectrode in a phototrophic microbial mat) but at least occasional assays for activity would be necessary to monitor the effects of and recovery from the perturbation. Of course, during and after recovery a microbial census can be done to determine changes in diversity or community composition.

The natural systems that have environmental ‘edge’ effects with likely changes in diversity are soils <pH 5 (Rousk et al., 2010) and hot spring mats where oxygenic photosynthesis wanes at temperatures above 65-70° C. Manipulating diversity within a natural sample can be a tricky business, as the extracted and manipulated community has to be reintroduced into a sterilized version of the system. This has been done employing a ‘model soil’ (Domeignoz-Horta et al., 2020). A laboratory-based model system has merits in terms of experimental control but certainly strays far from ‘realism’ and is dependent upon those microbes that have been cultured.

References

Domeignoz-Horta, LA et al. (2020) Microbial diversity drives carbon use efficiency in a model soil. Nature Comm 11: 3684

Donohue I et al. (2016) Navigating the complexity of ecological stability. Ecol Lett 19:1172-85. doi: 10.1111/ele.12648.

Li Z et al.(2025) A powerful but frequently overlooked role of thermodynamics in environmental microbiology: inspirations from anammox. Appl Environ Microbiol 91:e01668-24. https://doi.org/10.1128/aem.01668-24

Marquet PA et al. (2014) On Theory in Ecolog. BioScience, 64:701–710, https://doi.org/10.1093/biosci/biu098

May, R. M. Stability and complexity in model ecosystems (Princeton Univ. Press, 1973).

Prosser, JI et al. (2007) The role of ecological theory in microbial ecology

Nat Rev Microbiol 5: 384–392

Rousk J et al. (2010) Soil bacterial and fungal communities across a pH gradient in an arable soil. ISME J. 4:1340-51. doi: 10.1038/ismej.2010.58.

Scheiner, S. M., & Willig, M. R. (Eds.). (2011). The theory of ecology. University of Chicago Press.

I am going to take a hiatus from thinking philosophically about microbes for the next three weeks – we are travelling to Patagonia tomorrow. Most of the time will be on a ship going around Cape Horn and up the west coast of Chile admiring the fjords. I am looking forward to indulging my photographic instincts there …

Today’s Moment of Zen

I have lived in Maryland for about 6 months now, and it has been great seeing my family much more often and taking advantage of the arts scene here (sorry E Washington State!). However, I do get wistful about having those mountains so close by

Recent events in the U.S. public health sphere (withdrawal from WHO, halting publication of Center for Disease Control’s Morbidity and Mortality Weekly Report, and the feeble response to the proliferation of H5N1 influenza virus in dairy herds) stimulated me to look back at information I collected at the inception of the COVID pandemic. You remember those days – people disinfecting grocery purchases, making runs on bottled water and toilet paper, etc. At that period of uncertainty, I dove into the scientific literature to see what was known about another respiratory virus – influenza, thinking that some of that knowledge would be relevant to SARS-CoV-2. What I found shocked me – despite the influenza virus pandemic of 1918-1919 resulting in 20-100 million deaths and an annual 300-600 thousand deaths due to seasonal infections, we remain ignorant of many basic characteristics of its biology. This is what I call the “ecology” of the virus, used in the sense in which I said scientists studying microbial pathogenesis are ecologists – studying the interaction of an organism with its environment, in this case a human host.

I found a 2013 review of the routes of influenza transmission (Killingley and Nguyen-Van-Tam, 2013) – highly relevant to events in the U.S. in early 2020. Here is their list of important questions to resolve regarding influenza virus:

One hundred years on from a pandemic, many fundamental aspects of the infection process remained unknown. My general argument for fundamental work in microbial ecology is to elucidate principles that may have more general applicability to ecological problems, including applied ones. Here, my motives are specifically to advance ‘application’ – in particular, the advice proffered to government officials who had to implement public health policies with inadequate information. A specific example comes from #1 in the table – the U.S. adopted a very stringent 6-foot standard for social distancing for which there was no empirical data. Three feet would have been as effective and would have allowed more schools to be open.

Modeling

These fundamental elements of influenza transmission and infection (as well as data on viral replication kinetics at the cellular level) would be valuable to parameterize mathematical simulation models that could provide “what-if” projections on topics like social distancing, room capacity, and room ventilation. When I looked into what was the state-of-the-art for modeling, I found it to be barely advanced beyond the S-I-R compartment model (Susceptible-Infected-Resistant) developed about 100 years ago as 3 ordinary differential equations (Kermack and McKendrick, 1927). The team at Centre for Global Infectious Disease Analysis, UK Imperial College in London has gone beyond this with stochastic, spatially structured individual-based simulations of influenza and now SARS-CoV-2 (Imai et al., 2023).

Funding

In early 2020, I also looked into the nature of influenza research funded by U.S. agencies. It is unrealistic to expect the Centers for Disease Control to fund this work as its major mission is to support state and local public health agencies with information and data. Hence, I interrogated the National Institutes of Health research database. I found about $250 million in the budget for influenza – yet there was essentially no research regarding the epidemiology of virus infections or the kinetics of viral replication.

I revisited funding in January 2025 but focused on active grants for the NIH COVID-19 response. There are 460 projects, funded for a total of $3.5 billion. From titles and abstract searches, I was able to identify 3 projects that addressed viral replication kinetics at the cellular level and 2 on epidemiological modeling. These probably represented <1% of all COVID funding at NIH. The most visible foci of projects were vaccines, antivirals, and the immune responses to infection. These are all worthy targets but …

From my perspective as a microbial ecologist, the U.S. scientific establishment had not and is not funding the types of fundamental ‘ecological’ research that would have great value for scientists to provide advice on minimizing effects of the next epidemic/pandemic.

A wish list

I would have funding organizations promote much more research on transmissibility of respiratory pathogens. Relevant factors include:

• Infectivity and replication dynamics of the virus (and its changes over time as the pathogen evolves).

• How much virus is shed by an infected person over time and by different processes (breathing, talking, sneezing etc.)

• The quantitative importance of aerosol, droplet and contact transmission

• Environmental factors that impact virion survival in air-borne aerosols or droplets (for example, temperature and humidity)

• The indoor contact patterns of virus from an infected individual to susceptible ones, as a function of room size, occupancy, and ventilation.

Some of these can be relatively straightforward but others are rather difficult. I would advocate use ‘model system’ tissue culture cells to assay properties such as binding (to different cell types) and rates of virus release. In vitro systems have been designed to determine rates of virus inactivation under different temperature and humidity regimes.

Measuring virus release from infected persons involves more logistical and ethical issues, but there is a sampling device, appropriately called the “Geshundheit-II” (McDevitt et al., 2013).

The differences across time and by breathing activity would be invaluable, as would be adequate replication among individuals to develop a statistical distribution of these properties. This device has been used to measure SARS CoV-2 release by infected individuals (Tan et al., 2023). Also, at present it is unclear whether respiratory viruses are preferentially associated with a specific size class of particle, which would impact subsequent transmission dynamics.

Measuring the importance of different particle sizes in transmission and assessing risk from indoor contact patterns of infected and susceptible individuals is certainly more difficult and requires expertise beyond virologists to include engineers and physicists and physical chemists who have expertise in heterogeneous particle interactions and modeling of particle dynamics. See a review from physicists.

For example, not every respiratory particle emitted by an infected person will contain virions …

In general, modes of transmissibility have not been proven for respiratory viruses. However, there is a variety of circumstantial evidence that transmission of SARS-CoV-2 occurs primarily through aerosol particles with sizes ranging from 0.1 to 10 µm, and at relatively short distances (1-2 meters), but that mixing of infectious particles within an enclosed room may also contribute for longer-range transmission. Bathroom (especially communal ones) may also be a factor as SARS-CoV-2 is present in feces and can be aerosolized by toilet flushing [and per Dr. Germ, closing the lid before flushing doesn’t help.]. There are many opportunities for research projects that could firm up evidence regarding coronavirus, influenza virus and rhinovirus transmission and infectivity as viewed through the lens of the viruses’ “interaction with their environment” (aka ecology).

In particular, I envision that work of this type could lead to better parametrization of computational fluid dynamics models of classrooms, stores, restaurants on the risks of respiratory virus transmission under different conditions of crowding and ventilation. There would be improvements in advice to policy makers for indoor building codes and responses when the next epidemic/pandemic (inevitably) occurs.

On “classical” disease ecology

I have used a different take on the ecology of pathogens here – more frequently it encompasses studies of natural populations of animals that may act as intermediate hosts for human pathogens. That remains an important but understudied area. At the inception of the COVID-19 pandemic, this topic was scrutinized in regard to whether the origin of COVID-19 was from a market with wild animals or from a lab leak. That argument continues to this day

I don’t think we will ever know the origin with certainty. Both a “species jump” in virus host range and a lab leak are low (but nonzero) probability events – there should be global programs to control high-risk wet markets that sell live wild animals and tighten safety and maintenance at Biosafety Level III and IV labs around the world. Accidents do occur at the latter .

The interaction between bat species and coronaviruses is extremely interesting and complex. Bats are the richest source of diverse coronaviruses and have a remarkable capacity to host these viruses without overt signs of disease. Caves may host multiple species of bats that contain multiple coronaviruses. When these species form broods of young within similar time periods, colonies in this system may contain up to a million individual bats with high contact rates between mothers and young during early lactation (Hayman et al. 2013). As a consequence, individual cells of individual bats may be co-infected by different coronaviruses, generating loci of genetic reshuffling that could very well lead to ‘species jumps’ in host range.

This topic is quite far from my comfort zone — I would be happy to hear any comments from those with more expertise in this area.

Leave a comment

References

Anand, S., and Y. S. Mayya (2020) Size distribution of virus laden droplets from expiratory ejecta of infected subjects. Sci. Rep. 10, 21174.

Hayman DTS et al. (2013) Ecology of Zoonotic Infectious Diseases in Bats: Current Knowledge and Future Directions. Zoonoses and Public Health 60: 2–21. https://doi.org/10.1111/zph.12000

Imai, N et al. (2023) Quantifying the effect of delaying the second COVID-19 vaccine dose in England: a mathematical modelling study. The Lancet Public Health 8 (3), e174 - e183

Kermack W.O. and McKendrick, G.A. (1927). A contribution to the mathematical theory of epidemics. Proceedings of the Royal Society of London. Series A, 115(772): 700–721.

Killingley and Nguyen-Van-Tam (2013) Routes of influenza transmission. Influenza and Other Respiratory Viruses 7(Suppl. 2), 42–51. https://doi.org/10.1111/irv.12080

McDevitt JJ et al. (2013) Development and Performance Evaluation of an Exhaled-Breath BioaerosolCollector for Influenza Virus. Aerosol Science and Technology 47:444-451

Pöhlker, ML et al. (2023) Respiratory aerosols and droplets in the transmission of infectious diseases. Rev. Mod. Phys. 95, 045001. https://doi.org/10.1103/RevModPhys.95.045001

Tan, Kai Sen et al. (2023) SARS-CoV-2 Omicron variant shedding during respiratory activities. International Journal of Infectious Diseases 131: 19 – 25.

Today’s Moment of Zen

This photo taken in Korea struck me as particularly Zen-like, an attitude I am sorely in need of given recent events in the U.S.

Subscribe now

Subscribe now

The “science of science” is an emerging discipline that examines how and why scientific fields develop as they do. It tries to take into account not only the available methods and technology that determine advances in scientific disciplines but also historical, social, and economic factors. One approach is scientometrics, which analyzes “big data” from databases such as Web of Science, Scopus or US NSF Awards in order to catalog who and what gets funded and gets published, who collaborates with whom and career trajectories. A recent example is an analysis of the pedigree of faculty at US universities (Wapman et al., 2022). It found that 80% of tenure-track faculty received their degrees from just 20% of Ph.D.-granting institutions and that 8 institutions (Cal-Berkeley, Harvard, Illinois, Michigan, MIT, Texas, Stanford and Wisconsin-Madison) provided 20% of all domestically-trained faculty! The relationships were similar when broken down by discrete disciplines. [N.B. I do not mean to imply that they only ‘worthy’ outcome of obtaining a Ph.D. degree is becoming a tenure track faculty member. I think this paper speaks more to identifying a bias in that system. Understanding the bases for that requires further examination.]

Leave a comment

My analysis of microbial ecology through this lens will be shallower than much of the science-of-science literature, but represents my first step in this direction. I hope that I will be able to extend this in future posts, in particular to analyze how microbial ecology science is developing given its history, methods, and infrastructure in ways that might be different than other domains in the natural sciences and particularly its close cousins, microbiology and ecology.

The growth rate of microbial ecology

I have used Web of Science for most of my analysis, employing search terms to find published articles on microbial (Bacteria, Archaea, Algae and Fungi) ecology. Here is the increase in publications with hits from those search terms over the past 100 y. Fittingly for a microbial process, it is exponential – with a doubling time of about 8 years. This doubling time is similar to that found for all publications related to biology. The earliest papers concerned algae although I did notice a 1924 by EB Fred (microbiologist and later president at the University of Wisconsin) on the bacteria of Lake Mendota, a topic on which I worked some 50 years later!

Where is microbial ecology done?

Over the past five years, there have been approximately 30,000 papers published that mention microbial ecology according to Web of Science. These arise from 180 different countries. Note that more than 1/3 of these have at least one author from China and that 60% of the papers have an author from either China or the United States.

When it comes to specific institutions, analysis of affiliations is a bit more fraught as ‘collectives’ of academies, institutes and systems can dominate the list. I used the Web of Science “Affiliation with Department” filter – the top hits here tend to be Colleges or Faculties but do give some sense for very productive centers for microbial ecology research over the past 5 years.

The top end of microbial ecology research

Those 30,000 papers were published in about 3200 different publications. However, about 2000 of these were represented only once or twice and many of the remaining journals had aims very different than the philosophical basis for Think Like a Microbe (e.g., applied disciplines). My first idea was to restrict analysis to my so-called Top 5 journals (Ecology section of Applied & Environmental Microbiology, Environmental Microbiology, FEMS Microbiology Ecology, ISME Journal and Microbial Ecology). However, this represented only about 1500 publications which felt too restrictive. I then augmented my search with 11 other highly regarded journals associated with scientific societies to increase the yield to 3000, representing 10% of papers published over the past 5 years. Here are the top 50 affiliations of Colleges, Schools, and Departments with the departments highlighted in red to identify important foci of research.

It is quite pleasing to see that there are concentrations of microbial ecology talent as these are conducive to broad intellectually-stimulating interactions among both faculty and students which can produce novel ways of thinking about the fundamental problems in the field. I come at this from the perspective of my career in science – back in the day within a Microbiology Department if you weren’t working on Escherichia coli, Salmonella typhimurium or perhaps Bacillus subtilis, you were studying “funny bugs” and you probably were the only person in your department doing so. Your colleagues used “funny bugs” in the same pejorative sense that physicists call anything other than physics research “stamp collecting.” An important milestone in the United States was the Center for Microbial Ecology at Michigan State University, founded by Jim Tiedje in 1989 with funding from the National Science Foundation. In Germany, 2 Max Planck Institutes were begun in the early 1990s: Terrestrial Microbiology founded by Rudolph Thauer and Marine Microbiology initiated by Bo Barker Jorgensen and Fritz Widdel. These have also been exceptionally important to the development and growth of investigators who now are widely distributed contributors to the science of microbial ecology.

Leave a comment

Who funds research in microbial ecology?

My friend Jim Prosser has taken microbial ecologists to task several times for pursuing descriptive research rather than thinking more deeply to develop hypothesis-driven research. I agree with his analysis of the problem but would lay the problem not primarily with individuals but with funding agencies. Research scientists are certainly motivated by their passion for making progress on their field of interest, but they also want a successful career. In the academic world, that means publishing papers and in a field driven by experiments using advanced technologies that means securing and maintaining research funding. [I will note here this does not leave scientists (particularly senior scientists) off the hook, as they are often intimately involved in setting agency objectives and determining which projects get funded].

For the 3000 papers published in last 5 years from my set of 16 top journals, here are the funding sources that contributed to at least 2% of those papers.

The list bears some relationship to the order of countries by number of publications. China’s NSF and the 3 major US government funding agencies (NSF, NIH, and DOE) comprise nearly 50% of the funding sources. Of note is that 2 of the entries are not government funding agencies – the Moore Foundation and the Simons Foundation. Both have been active in funding marine microbial ecology projects.

What gets funded and why

The first question can be objectively answered from databases. I downloaded information on all current projects from the US National Science Foundation’s Ecosystem Sciences Program and Population & Community Ecology Program. Within the U.S. funding structure, these are the 2 programs I would expect to be best aligned with the philosophy of Think Like a Microbe: identifying principles that have a reasonable degree of generality. From a concatenated list of abstracts (about 300 from each program – note that this was not restricted to just projects involving microbes), I extracted word frequencies which I then culled to those I thought were most representative of the topic or intellectual activity. Rather than a table, I regressed to an early 21^st C convention and made a word cloud.

Ecosystem Science

Population and Community Ecology

The second question of why something get funded is more difficult and more of a sociological one. My experiences were with the U.S. funding system, primarily the National Science Foundation (NSF) and the Department of Energy (DOE) – a caveat is that my reflections are somewhat dated (as I am now 10 years out of the game). However, I hope my insights are of use to some young scientists. Dealing with the nuances of funding agencies is generally not part of one’s academic training. In the words of Omar Little, that great philosopher from the television series, The Wire, “The game is out there, and it's either play or get played.”

NSF is traditionally thought of as the “basic science” agency in which projects are driven by curiosity and discovery. DOE fashions itself as more “mission oriented” with respect to looking for science and technology solutions to energy and environmental problems. In reality, one can find “mission-oriented” elements in NSF and quite basic research projects in DOE.

It is also relevant to note the outsized impact of DOE’s Joint Genome Institute on the shift of microbial ecology research to DNA sequence-based projects. JGI was established in 1997 to work on the Human Genome Project. When that project concluded, the sequencing machines (120 at maximum) and workforce remained so DOE searched for other uses. In 2004 JGI started its Community Sequencing Program where investigators could receive large-scale sequences of genomes or metagenomes that they had sampled. In addition, DOE’s Office of Biological and Environmental Research started its Genomes to Life program in 2002 to explicitly provide research project funding for those proposing DNA-sequence based analyses of microbial ecosystems.

There is, I believe, important cultural difference between these agencies in how projects are chosen for funding. I served on review panels for each. In the case of NSF, the advisory panel is dominant. At the end of the review, the panel constructed a ranked list of projects’ priority for funding and the program officers did not deviate from that ranking. At DOE review panels, the programs officers made it quite clear at the outset that the role of the panel was “advisory” and that in the end the officers would be making the decisions on which projects to fund. In practice, this meant the officers would take into account how well “aligned” the projects were with current thinking of mission priorities (and these could change over 5 year periods). At the time I served on the NSF Ecology panel, a culturally important criterion for the panel was that the project addressed an important (theoretical) question in (macrobe) ecology. I was the only microbial ecologist on the panel, and most microbial ecologists submitting proposals at that time had a relatively weak formal training in ecology so this aspect rarely shone through in their proposals.

A few last thoughts:

• Program officers are generally happy to talk to you about their program and how to best structure your proposal

• It never hurts to have an experienced colleague read over your proposal and provide suggestions that might increase its “fundability.”

• Persistence is necessary – NSF reports that the funding rate for Division of Environmental Biology was 25% in 2021, but I am highly skeptical of this; the percentage of grants that were likely to be funded from my review panels was no more than 10%. Hence, one cannot get discouraged by rejections – it has often struck me that success in research science has less to do with how ‘bright’ you are than how persistent you are.

• The U.S. system is kind of a lottery, given these proposal success rates. So you may need multiple rounds of submission, altered versions of a central idea that can be submitted to more than one agency or program, and several different lines of research for which you can submit proposals.

Some science-of-science investigators have brought attention to whether grant funding systems favor safe research over ‘risky’ proposals (where risky is used in the speculative sense – there can be high reward but also a chance of no progress). They generally conclude this is a problem, as grant agency officers, review panelists, and principal investigators themselves act in risk-averse ways and reviewers use bibliometrics (publications and citations) as a crutch rather than in-depth analysis of novelty and merit. A particularly dramatic example of this is Dr. Katalin Kariko, who ‘failed’ in academic research but went on to develop mRNA-based vaccine against the COVID virus.

See: https://www.journals.uchicago.edu/doi/10.1086/719252

Reference

Wapman, K.H., Zhang, S., Clauset, A. et al. Quantifying hierarchy and dynamics in US faculty hiring and retention. Nature610, 120–127 (2022).https://doi.org/10.1038/s41586-022-05222-x

Your moment of Zen

The mouth of the Tagus River, Lisbon, Portugal

Back in an earlier geological age, ecology was neatly divided into autecology and synecology. Synonyms for these terms are physiological ecology and community ecology. The more recent theoretical developments of ecosystem ecology (and other larger-scale analyses) has tended to also be placed in synecology .

Within microbial ecology, progress in physiological ecology has been rather fallow in the 21^st C (with an exception noted below). I readily admit to a personal bias in its favor — it is what first attracted me to microbial ecology. For Clements and Shelford (1939), “(plant) ecology is physiology carried into the actual habitat, and in consequence its paramount theme is stimulus and response.” Some waggish grad students working on microbial mats in Yellowstone National Park under Tom Brock paraphrased this as “ecology is physiology under the worst possible conditions.” The theme of stimulus and response spoke strongly to me. So did Arthur Koch’s thoughts regarding the macroeconomics of microbial growth – in most environments, growth rate is limited by the concentration of some essential resource whether by supply or due to microbial proliferation at its expense. Resource supply rates may be variable over time. Microbes’ adaptations to these problems (that can differ among species and in different habitats) speak to the nature of the environmental fluctuations that have driven their natural selection and hence can generate principles that may apply widely (even if not universally) in microbial ecosystems.

For example, when I taught a Microbial Ecology course, I provided students with a summary ‘cheat sheet’ of some strategies we had discussed in depth, with respect to resource acquisition.

Leave a comment

The notion of using “physiological indicators” in the field to ascertain growth rate-limiting resources was most extensively developed for phytoplankton. These included elemental compositional analysis (ratios of C, N, and P or specific determination of protein, storage carbohydrates and RNA) or analysis of uptake rate (V_max normalized to biomass) for P, N, or light. Although these approaches have fallen out of favor and supplanted by Big Data analyses of DNA sequences, ironically the ‘omics revolution does provide an improvement on them – transcriptomics. I have been encouraged by the substantial increase in the number of publications in the Big 5 microbial ecology journals that have featured this approach. This does require well annotated genome sequences of the relevant species. However, note that analysis after a perturbation to the system may identify not only ‘usual suspect’ genes but also those of unknown function, thereby providing initial insight into their role in the organism’s ecology.

As alluded to above, there is a sub-field of physiological ecology that has had significant momentum over the past decades, microbial pathogenesis. I understand that scientists working on pathogenesis may not perceive themselves as microbial ecologists but they are in fact working on the “interaction of microbes with their environment.” It represents one of the four microbial-scale ecosystems I identified in my initial post of this substack, Macroorganism-associated, with its capacity for coevolution and sophisticated surface and (in some cases) intracellular interactions.  Consider Salmonella enterica serovar Typhimurium which uses type III secretion systems to inject up to 40 distinct proteins into host epithelial cells, as a prelude to invasion and replication of Salmonella cells within the host cell. Or the ‘games’ played by animal hosts and some pathogens with respect to the essential nutrient, iron. One line of host defense against infection is to sequester transferrin-bound iron in certain organs to prevent its acquisition by pathogens. Pathogen microbes may elaborate high-affinity siderophores to out-compete the host for iron in circulation (Nairza and Weiss, 2020).

Certainly investigations into such host-pathogen systems have less “complexity” in that the researchers are generally interested in one specific microbe. Furthermore, a number of frank pathogens can be experimentally manipulated using genetic techniques of mutagenesis to identify virulence factors (such as S. enterica’s injected proteins) and comparisons made from the pangenome of multiple strains. These advantages are less frequently found among abundant lineages of bacteria from aquatic or soil habitats.

A valuable counterpoint to that has been the success in analyzing the physiological ecology of two abundant lineages in the ocean – representatives of SAR11 and Prochlorococcus. These clades typically have a small cell volume and genome size. Furthermore, transcriptional regulation in representatives of these clades is diminished such that most genes are expressed constitutively (Noelle et al). Understanding the selective forces that have driven them to these physiological characteristics is relevant to the mechanisms that drive ecological dynamics in aquatic ecosystems.

This work can be construed as studies of “model organisms” for that particular habitat. The study of organisms such as Drosophila melanogaster and Caenorhabditis elegans has been extraordinarily fruitful [pun intended] across a much broader breadth of biology than just those species. There is no reason to think it is otherwise in the realm of microbial ecology. DNA-based censuses of other environments suggest taxa that are highly abundant and hence whose physiological ecology may be representative of other microbial ecosystems. Prevotella copri is very abundant in the human gastrointestinal tract and a culture has been isolated (although I find only 3 papers focused on its properties in culture). Verrucomicrobia and Acidobacteria are two of the most abundant phyla found in soil. Yes, they are underrepresented among cultured taxa but with concentrated effort taxa from these clades have been cultivated.

As mentioned above, transcriptomics provides a particularly attractive tool to assess an important aspect of the physiological responses made by model organisms to relevant environmental conditions. For example, Scales et al. (2023) interrogated a set of nine Curtobacterium strains isolated from soils differing in climate aridity by challenging them with dessication. They identified differences in survivorship among strains as well as up-regulation of genes for up-regulation of genes involved in DNA damage repair, oxygen scavenging, chaperonins, osmolyte production, and efflux pumps. The technology is broadly available to address issues in physiological ecology by these systems-biology approaches.

References

Clements, F.E., and Shelford V. 1939. Bio-ecology. New-York: Chapman & Hall.

Huang F et al. 2021. Cultivation of the gut bacterium Prevotella copri DSM 18205^T using glucose and xylose as carbon sources

Microbiology Open 10: e1213. https://doi.org/10.1002/mbo3.1213

Koch AL. 1985. The macroeconomics of bacterial growth. In Bacteria in their Natural Environments (M Fletcher and GD Floodgate, ed.) pp. 1-42

Nairza M and Weiss, G. 2020. Iron in infection and immunity. Molecular Aspects of Medicine 75: 100864. https://doi.org/10.1016/j.mam.2020.100864

Scales NC et al. 2023. Desiccation induces varied responses within a soil bacterial genus. Env. Microbiol. 25: 3075-3086. https://doi.org/10.1111/1462-2920.16494

Leave a comment

Today’s Moment of Zen

The Coast Guard Station on Whiskey Island, Cleveland OH