Tell me about a complicated man seed.

— Homer’s Odyssey, translated by Emily Wilson (revised by Aaron Suiter)1

Michigan lily (Lilium michiganense) has two modes of reproduction: sexual (through seeds) and vegetative (through bulb offsets, which are clones of the mother plant). In the nursery setting, it is reportedly faster to propagate Michigan lily with bulb offsets, but I prefer to grow them from seed. Either method will eventually produce a full-sized bulb (see figure 1).

From the plant’s perspective, neither reproductive strategy is inherently better. For an individual Michigan lily that has found itself in an environment favorable to its growth, vegetative reproduction makes sense: it can use the stored carbohydrates in the bulb offset to establish quickly, giving it the edge it needs to grow tall enough to avoid being shaded by neighboring vegetation.

But vegetative reproduction creates genetically identical clones, which could diminish the population’s resilience to stressors. If pest pressure increases, climatic conditions change, or moisture levels rise or fall, genetic diversity could make it more likely that some portion of the population survives and reestablishes. Plus sexual reproduction allows plants to send offspring further away, where conditions might be more conducive to survival.

I’m a seed ecology blogger, so naturally I prefer growing Michigan lily from seed. And I’d rather my plants have as diverse a gene pool as possible, which in theory could help populations survive adverse conditions (e.g., pest/disease pressure, pollution, suboptimal climatic conditions).

How to germinate Michigan lily seeds

Although very little has been published on propagating Michigan lily, there is a fair amount of information out there on propagating the closely related Canada lily (Lilium canadense), and Peter Heus published a comprehensive protocol for that species in Native Plants Journal.2 Michigan lily and Canada lily are so similar in appearance that, back in the 1950s, Boivin and Cody proposed that L. michiganense be classified as a subspecies of L. canadense.3 Based on my experiments, Michigan lily has deep simple epicotyl morphophysiological dormancy, just like Canada lily.4 So to germinate Michigan lily, you can basically just apply Heus’s protocol. With that being said, I’ve conducted extensive germination experiments on Michigan lily to determine what stratification regime yields the best germination, so the protocol I present below is more optimized for Michigan lily.

I won’t go into the details of deep simple epicotyl morphophysiological dormancy here (that’s the topic of a future post), but there are two key things to know about this dormancy type:

• Seeds of this dormancy type need to experience warm–moist conditions followed by cold–moist conditions to germinate.

• The shoot (epicotyl) has additional dormancy beyond the root (radicle). So the root will emerge after warm stratification, but the shoot will not emerge until after cold–moist stratification.

In my experiment, Michigan lily seeds achieved the best germination when given this four-stage protocol:

1. 6–12 weeks simulated summer (warm–moist stratification with 30°C days and 15°C nights).

2. 4–10 weeks simulated autumn (warm–moist stratification with 20°C days and 10°C nights). During this period the radicle (root) will emerge from the seed, and a bulblet will form.

3. 8–12 weeks simulated winter (cold–moist stratification). This period releases dormancy of the shoot (epicotyl).

4. Germination completes when seeds are placed in a warm environment. I’ve only tested simulated spring temperatures (20/10°C), but I’m guessing other warm temperatures would work.

Below I explain all of that in more depth.

Inducing radicle (root) emergence in Michigan lily

Michigan lily seeds have a multi-part germination process. Warm–moist stratification induces the radicle (root) to emerge (see figure 3), but the epicotyl (shoot) remains dormant and will not emerge until after an additional period of cold stratification. There are two regimes you could use to warm–moist stratify the seeds: one is simpler (but yields lower radicle emergence) and the other is more complicated (but yields higher radicle emergence).

The simpler method (lower radicle emergence)

Mix together equal parts of perlite and vermiculite and then moisten it. Just as in cold–moist stratification, you want the substrate damp, but not sopping wet. Then mix in the seeds and let the seeds incubate at 20/10°C for 16 weeks. I like to put the seed/perlite/vermiculite mixture into mason jars or tupperware, with at least a couple inches of substrate to accommodate root growth. In my experiment, this resulted in 37% radicle emergence. Not amazing, but seed is cheap.

I realize not all readers have a setup for incubating at an alternating temperature regime. You have two options: attempt to incubate them at room temperature (which is approximately 20°C), or else incubate them outside in fall. Heus’s protocol for Canada lily is 17 weeks of constant 21–23°C, so I’m guessing that method would yield reasonable germination for Michigan lily. If you opt to stratify outside, just make sure the seeds are in the shade, or else direct sun will heat up your seeds and cook them. I incubate in darkness (I wrap my vessels in aluminum foil), but that may not make a difference. Better yet, if you stratify outside, start them in the summer so that they get both summer temperatures and fall temperatures (more on this below).

The more complicated method (higher radicle emergence)

Based on my experiments, Michigan lily seeds have highest radicle emergence when they experience summer conditions (30/15°C) followed by fall conditions (20/10°C). I tested 9 different 16-week protocols with varying fractions of summer and fall conditions. Radicle emergence was highest for seeds receiving 6–12 weeks of summer conditions followed by 4–10 weeks of fall conditions (see figure 4). 8 weeks of 30/15°C followed by 8 weeks of 20/10°C had the highest overall radicle emergence, but it did not differ significantly from the 6w / 10w, 10w / 6w, and 12w / 4w treatments.

So if you’re trying to maximize radicle emergence (and thus overall germination), my recommendation is to use 8 weeks of summer temperatures followed by 8 weeks of autumn temperatures. As with the simpler method, I use equal parts of perlite and vermiculite as my substrate, and incubate in darkness.

Inducing epicotyl (shoot) emergence in Michigan lily

Michigan lily seeds have epicotyl dormancy, meaning that the shoot remains dormant even after the root is produced. Fortunately, breaking the epicotyl dormancy is easy: just put the seeds in the fridge after root emergence. I haven’t yet experimented on the minimum time required to break the shoot dormancy, but 8 to 12 weeks should suffice. If you leave them in too long, the shoots come up in the fridge, but that’s not a big deal.

So when I do this, I just put my mason jars or tupperware (filled with vermiculite/perlite and seeds with roots) into the fridge. After 8 to 12 weeks, I pour out the mixture on top of prepared pots filled with PRO-MIX potting media. A few weeks later, the shoots emerge, and often the seedlings only grow a single leaf in their first year (see figure 5).

Raising Michigan lily seedlings to transplantable size

Michigan lilies have what I call “type 2 growth,” meaning that they have a tendency to stop growing and senesce, even when conditions are ideal for growth. In my experience, many seedlings produce nothing more than a single leaf in their first season, and then senesce (see figure 6).

After the seedlings have begun to senesce, I move them to my fridge to release their dormancy, a process I describe in more detail here. I haven’t yet determined the minimum duration of chilling needed to restart growth, but 2–3 months is a safe starting point. By using my fridge to chill them, I can compress a year-long growth cycle into about 7 months, enabling me to grow these faster. After removing the seedlings from the fridge, they emerge as shockingly large seedlings (see figure 7).

In their second season, the lilies are still too small to flower. You can proceed one of two ways: either plant the second-year seedlings out and let them finish maturing outside, or else continue cycling them through periods of cold until you’re satisfied. Planting them outside is easier, but once they’re in the ground, they’re doing one season of growth per year. If you keep them in pots, you can induce about two seasons of growth per year with the artificial chilling method.■

Have a question? If you have questions about Midwestern native plant propagation that you want answered in a future post, I’d love to hear from you! Shoot me an email at plantpropagationproject@gmail.com

Want to support this work? Buy me a cup of coffee (or a bag of potting medium)!

Buy me a coffee!

Want to stay up to date on all my propagation protocols? Subscribe for free to receive new posts and support Plant Propagation Project:

Subscribe now

I have only rarely seen purple milkweed in the wild, which can be explained by the fact that it is typically found in oak savanna, an ecosystem that has been virtually eliminated from Wisconsin. On a more hopeful note, there is at least one report of purple milkweed reappearing when degraded oak savanna was restored.1 So if you happen to have a degraded oak savanna, clearing and burning it might just bring you purple milkweeds.

In addition to being rare in the wild, purple milkweed is rare in cultivation; very few nurseries sell this plant, and those that do only sell limited quantities. Seed is also challenging to buy, and as of this writing, no reputable native seed seller that I know of has the seed in stock.2 Since plants are often slightly easier to come by (Prairie Moon tends to offer them, briefly, every year), if you want to grow this plant from seed, it might be easiest to first establish a small population from purchased live plants, and then collect seed from those plants.

In his 2009 article “Ecology and Conservation of Purple Milkweed” (available to read online with a free JSTOR account), Thomas Brock reports extensively on his work growing purple milkweed from seed, but I feel the need to contribute my own experiences to the discussion of growing this plant, since I’ve had very different results than Brock. Brock, for instance, reports extremely high germination percentages for purple milkweed: 50 out of 50 seeds germinated for one lot collected in 2005, and 46 out of 50 germinated for another lot collected in 2007.3 But while he saw germination as high as 100%, I never saw germination higher than 9% in my own trials. And while he generally saw good survival when transplanting into the field, I’ve had mixed results: one year I saw 100% survival in my transplants (with five out of six blooming their first year), but this year most of my transplants have died (and none have flowered).

So assuming that you’ve gotten your hands on the much-coveted purple milkweed seed, you’re probably feeling the pressure to not mess up growing these plants, and I’m hoping the information below is helpful. Like many of my posts, this is a work in progress, and I’ll update this post as new results come in.

My Experience Germinating Purple Milkweed Seeds

I collected my purple milkweed seeds on October 11, 2025, then stored them dry, at room temperature, until January 15, 2026. I then divided the seeds into two groups, each consisting of 45 seeds, and subjected one group to 6 weeks of cold-moist stratification and another group to 12 weeks of cold-moist stratification. After pretreatment, I sowed the seeds into ProMix-filled pots, lightly covered with vermiculite, and germinated the seeds at simulated spring temperatures (20°C days, 10°C nights). For more information on the temperature regimes I use, see my article on that topic.

In both groups, germination could be observed as early as day 5, and total germination was assessed at day 14. All seeds that germinated developed quickly into robust seedlings (see figure 2).

I observed the seeds until day 28, but no additional germination occurred after day 14. The group stratified for 6 weeks had 1 seed germinate, and the group stratified for 12 weeks had 4 seeds germinate (see table 1). The difference in germination between the two groups was not statistically significant, meaning that we cannot rule out that the difference observed was due to random chance. Notably, my germination percentages are much, much lower than Brock’s. It’s possible that this could be due to minor differences in methodology (e.g., his seeds were fully exposed to light, whereas mine were dusted with vermiculite), but my hunch is that my seeds had lower viability than his. Perhaps in future years I’ll systematically test viability, but based on vibes alone, I can say that the seeds I collected lacked the plumpness that I like to see in Asclepias seeds.

It’s also worth noting that I didn’t test germination for seeds that were not stratified at all, despite the fact that many milkweeds do not require or benefit from stratification, and in future years I’ll test an unstratified control.

So, provisionally, my advice is to stratify for about three months, but I’ll update that advice when I achieve better germination.

Growing Purple Milkweed Seedlings in Containers

Some milkweeds, such as whorled milkweed (Asclepias verticillata), languish when seedlings are grown at simulated spring temperatures (20°C days, 10°C nights), and I therefore grow most of my milkweeds at simulated summer temperatures (30°C days, 15°C nights). Purple milkweed, however, seemed perfectly happy to grow at simulated spring temperatures, which makes sense given that it grows in soils kept cooler by dappled savanna shade.

The main thing to know about growing purple milkweed in containers is that it grows really fast. In my experience, purple milkweed seedlings fill out 1-quart pots in about 7 weeks (see figure 3), and they fill 2.5-in.×2.5-in.×3.5-in. pots even quicker. My impression (unsubstantiated by any rigorous testing) is that they don’t like being transplanted too early in the season, so I recommend sowing them in late April or early May so that they’re ready to plant out into warmer spring conditions in mid-June.

As with all my containerized seedlings, I fertilize regularly. Brock notes that, in the absence of fertilizer, seedlings become chlorotic,4 so fertilization may be essential to seedling survival.

Planting Out Purple Milkweed Seedlings

My experience of planting purple milkweeds out is that they are prone to spontaneously dying, and I have no idea why. I planted some out in early May, and those quickly senesced (see figure 4). I initially assumed they were sad about colder conditions or shorter day lengths, but I’m no longer confident in that. I planted more seedlings in mid-June, when days were long and weather was warm, and many of those died too. The specimens that survived were also more shaded by surrounding shrubs, so it’s possible that purple milkweed needs at least some shade to survive transplanting. With that being said, last year I transplanted all my purple milkweeds into full sun, and 100% survived, though they did get shaded a fair amount by aggressive pokeweed (Phytolacca americana) later in the season.

Next year I’m planning to run field trials testing whether various levels of shade (achieved using shade cloth) affect seedling survival. Stay tuned for that.

A Note About Purple Milkweed Morphology

Purple milkweed specimens, even those belonging to the same population, can take on a wide variety of vegetative forms (see figure 5). It is therefore very difficult to identify a purple milkweed without seeing its flower (or seed pod).

In his article about purple milkweed, Brock emphasizes this point, writing:

If the plant is not flowering, it cannot be positively identified as purple milkweed. Although field guides often confidently describe vegetative characteristics, most of these are variable and uncertain. Frequently they are based on the characteristics of a few herbarium specimens. According to Farnsworth and DiGregorio (2001), misidentification in the field has led to incorrect labels on at least one voucher specimen at the Gray Herbarium at Harvard University.5

So don’t try to identify your purple milkweeds based on vegetative characteristics alone! ■

Have a question? If you have questions about Midwestern native plant propagation that you want answered in a future post, I’d love to hear from you! Shoot me an email at plantpropagationproject@gmail.com

Want to support this work? Buy me a cup of coffee (or a bag of potting medium)!

Buy me a coffee!

Before I begin the technical portion of this article, it’s worth dwelling for a moment on the word stratification. In any other context, the word stratification would carry a sense of layers (i.e., strata): take—for example—geology, where it refers to the process by which layers of rock are formed; or sociology, where it refers to the hierarchical layering of society by class, status, and power; or music, where it refers to the layering of musical texture. So why do we use this word to describe cold-moist pretreatment of seeds?

As is often the case, the answer is history. The earliest usage I could find of the word stratification in the context of pretreating seeds in damp media comes from Robert Thompson’s 1859 publication The Gardener’s Assistant, in which he describes stratification as a process “which consists in placing a layer of seeds on the surface of the ground, then a layer of sand or light soil from 1 to 2 inches in thickness, then another layer of seeds, and so on.”1 So in this early usage, the process of pretreating is in fact a process of layering. To complicate matters, Thompson clarifies that “it is immaterial whether the seeds are placed in alternate layers, or simply mixed with the soil.”2 So why call it “stratification” at all, if the layers are immaterial to the process? Thompson makes no claim to have coined the term, so it’s likely he’s just reporting a term already common in horticultural circles. At any rate, 150 years later, the process of stratification no longer involves layers, but “stratification” is the word we’re stuck with.

How to Stratify Seeds for Native Plant Propagation

Materials

Assuming you already have a refrigerator, you only need two things to stratify seed:

• Plastic bags. Any small reclosable plastic bag will work. For any given species, I tend to stratify about one-packet-worth of seeds, so I favor 2-in. × 2-in. plastic bags (see figure 1). So for folks buying seed packets from vendors like Prairie Moon—or folks trading packets with friends—this is a great size. If you’re doing larger quantities than that, you’ll need bigger bags.

• Play sand. Just buy play sand at your local hardware store. If you have multiple play sand options, choose the option with the smaller particle size. The sand I use has a very small particle size, which I can best describe as the type of sand you would expect at a luxury tropical resort (see figure 2). Ultimately the seeds don’t care that much how fine your sand is, but for my exact process, fine sand is better. Just make sure it’s not something gritty like paver’s sand. I would strongly advise against trying to collect your own sand; the play sand sold at hardware stores is washed, and is unlikely to introduce pathogens or organic material to your seeds.

Process

Step 1: Measure out your sand.

I like to use a ratio of at least 4 parts sand for every 1 part seed (see figure 3). So if you have 1 teaspoon of seed, use approximately 4 teaspoons of dry sand, or more. There’s no major disadvantage to using too much sand, as long as you’re being reasonable (don’t use a 100:1 sand–seed ration). Precision does not matter here, this is just a rough guide. You should aim to use at least 1 teaspoon of sand, no matter how little seed you have. So if you’re working with tiny seeds, even if you only have ⅛ teaspoon of seed, still use 1 teaspoon of sand.

To be totally transparent, I never measure my sand. I simply whip up a big batch of damp sand, and then take scoops of it for each species of seed I’m growing. The measurements presented here are intended to give beginners a solid footing to stand on.

Step 2: Moisten your sand.

It’s now time to moisten your sand. In figure 4, I show seven levels of hydration ranging from bone dry (A) to sopping wet (G). For each hydration level, I show what it looks like molded into a dome (along the top), and also what it looks like loose (along the bottom). My recommendation is to aim for something like D or E. But if you used something more like C or F, your seeds probably won’t care.

As your seeds absorb moisture from the sand, the sand will dry out slightly. So the risk of using a hydration level comparable to B or C is that, after a few weeks, it may be bone dry like A. But if you start with sand at hydration level E, even after the seeds fully hydrate, the sand will be at level C or D, which is sufficient.

Step 3: Mix in your seeds.

Nothing hard here, just mix your damp sand with your dry seeds (see figure 5).

Step 4: Bag your seed–sand mixture.

Now put the mixture in a plastic bag (figure 6). Make sure to label your bag!

Step 5: Refrigerate your seed–sand mixture.

We’re aiming for a temperature of approximately 40°F (4°C), which is the standard fridge temperature. A little higher or lower is fine, but avoid temperatures below freezing. Especially for first-time stratifiers, check your baggies after about a week to make sure the moisture level looks correct. If you’re seeing dry sand, add water (just a drop or two at a time).

As for timing, know that there is no perfect source of information on ideal stratification durations. Prairie Moon is perhaps the most comprehensive source of recommended stratification durations in that they list a time for every species they sell, but their information is not always accurate. Generally speaking, for plants that require only cold stratification to germinate, 8 weeks is sufficient to get a reasonable amount of germination. Sometimes germination peaks at some point after 8 weeks, and sometimes it peaks before 8 weeks—it depends on the species. But if you’re at a loss for timing, just try 8 weeks and see what happens. Or search Google Scholar for peer-reviewed protocols. Or leave a comment on this post and I’ll do my best to answer!

If you’ve not grown a particular species before, and are not sure how much time in the fridge it requires, check your baggies every few weeks to inspect for germinating seeds. If seeds have germinated in their bags, they are still totally fine to sow, but the longer you leave germinated seeds in their bags, the higher your risk that the germinants start to rot (and they also get more difficult to work with as they get larger).

Step 6: Sow your seeds.

After your seeds have been sufficiently pretreated, it’s time to sow them. I don’t like potting up, so I usually sow my plants in whatever container I plan to grow them to maturity in. Gently tap out a small aliquot of sand–seed mixture into each container (see figure 7). Depending on the species, I aim for about 3 to 10 seeds in each pot or plug.

I then use a spray bottle to mist the sand–seed mixture such that the sand washes off the seeds and disappears into the substrate, ensuring good seed–substrate contact and allowing me to inspect how many seeds are present (see figure 8). At this point, you’re ready to cover the seeds with vermiculite/grit/peat (or not, depending on the species) and set the container in your germination area. Generally speaking, if the seeds were viable and received a suitable length of pretreatment, you should start seeing sprouts within about 2 weeks.

That’s all! In future posts I’ll explore some common stratification mishaps, but for now, just leave a comment if something has gone terribly awry with your seeds. ■

Have a question? If you have questions about Midwestern native plant propagation that you want answered in a future post, I’d love to hear from you! Shoot me an email at plantpropagationproject@gmail.com

Want to support this work? Buy me a cup of coffee (or a bag of potting medium)!

Buy me a coffee!

Start Survey

Prairie Moon claims that prairie lily (Lilium philadelphicum)—also known as “wood lily”—takes five years to mature and flower.1 Well it’s been seventeen months (520 days, to be exact) since I sowed my prairie lily seeds, and I have my first prairie lily flower (see figure 1). So at least in my carefully controlled growing conditions, prairie lily can be raised to flowering age in well under five years.2 For all the readers out there who are interested in growing this beautiful native lily, but are daunted by the thought of waiting half a decade, I’m here to tell you that you don’t actually have to wait that long!

Prairie lily was once abundant in the mesic prairies of south-central Wisconsin, but only about 100 acres of the original 800,000 acres of mesic prairie remain in Wisconsin, and with that habitat loss, prairie lily has become much rarer on the landscape. Many prairie restorations in Wisconsin are on mesic soils (typically farmland converted to prairie), but I don’t see many prairie lilies on these restorations. Although I don’t know of any formal research on individual or population survival of prairie lily in restorations, there are a variety of reasons that prairie lilies may struggle to survive:

• Animals preferentially dig up and eat their delicious bulbs

• Animals preferentially browse their delicate stems and leaves

• Taller, more aggressive plants outcompete and smother them

• They have relatively short lifespans, so need to successfully reproduce frequently to maintain a population

When it comes to growing prairie lily from seed, germination is easy and straightforward, but ensuring their survival to flowering age is a little trickier. I’ll briefly describe my experiences germinating these seeds, and then share my thoughts on getting them to flower.

How to Germinate Prairie Lily Seeds

Prairie lily is very easy to germinate; my protocol yields germination rates of about 90% and requires only nine weeks of cold stratification (and I bet you could get away with less time in the fridge).

Although I typically use a 20/10°C temperature regime when sowing these seeds (read more about temperature regimes here), prairie lily also germinates well at room temperature (approximately 20°C), which makes it a good choice for those who don’t want to mess around with precision temperature control. I’ve also tested its germination with a 15/6°C temperature regime, and I did get some germination, but it was much slower (see table 1).

How to Ensure Your Prairie Lily Seedlings Reach Maturity and Flower

I strongly advise against planting out prairie lilies that are less than eight inches tall (a height they typically achieve after one dormancy cycle).3 I’ve tried planting ten-week seedlings (which were mostly single cotyledons, or just a few basal leaves), and they all died. I’ve also tried planting out five-month seedlings (see figure 3), and those also died. So patience is key; do not rush to plant these out! Young plants are very small, and highly prone to being smothered by their neighbors.

I grow my prairie lilies indoors until the foliage senesces, then expose them to one cycle of artificial chilling.4 This chilling cycle transforms very small seedlings comprising only basal leaves into caulescent plants (see figure 4). I’ve misplaced my notes from last year, so I can’t tell you exactly what the timing looked like. My guess is that it took about eight months for the seedlings to senesce after germination, and then I moved them to the fridge at ten months (only because I was neglectful). I know for sure that I removed them from the fridge at fifteen months, and then planted them into my garden at sixteen months.5 I waited a month to plant them because I wanted the plants to be tall enough that I could see them in the garden, lest I accidentally dig them up or step on them.

Out of about twenty-five plants, only one flowered so far this year. So a second dormancy cycle is likely necessary to achieve high bloom rates. I’m just going to let the lilies experience this second chill cycle naturally (in the ground) this winter, but you could also keep them in their containers and do another artificial chill. I opted not to do this because it would require potting them up, and it would take up quite a lot of fridge space.

When planting into the garden (or into restorations), I recommend encircling them with a cylinder of hardware cloth mesh, which protects their delicate stems from herbivores and humans. Also be aware that they have very floppy stems, so plant them amidst vegetation that can help prop them up. ■

Have a question? If you have questions about Midwestern native plant propagation that you want answered in a future post, I’d love to hear from you! Shoot me an email at plantpropagationproject@gmail.com

Want to support this work? Buy me a cup of coffee (or a bag of potting medium)!

Buy me a coffee!

Live plants of downy gentian are very rare in the nursery trade (the notable exception being Windflower Natives), and it is also very difficult to find seed for this plant on the market. Prairie Moon offers small quantities of the seeds sporadically (over the last 3 years, I think they’ve offered seeds once), and that’s about it. Here in Wisconsin—and perhaps elsewhere in its native range—seed pods often fall prey to moth larvae, and they are therefore often found filled entirely with frass rather than seed.2 For this reason, restorationists struggle to collect enough to establish new populations of this special plant.

Because downy gentian seed is so scarce—and thus precious—downy gentian is a good candidate for container culture, where every viable seed can be nursed to maturity.3 If you can only collect a few hundred seeds in a year, you’d probably rather ensure that each of those seeds yields a plant rather than scattering them outside and hoping the birds and insects and fungi don’t take them. The good news is that, like other members of the Gentiana genus, downy gentian seed pods contain many, many seeds, so a single frass-free seed pod can yield a hundred or more plants (though for the sake of genetic diversity, more pods would be ideal).

The bad news is that, if you’re a home grower, you may struggle to find downy gentian seeds on the market. I think that there is opportunity to establish populations of this plant for seed production, and to responsibly manage larval infestations in a way that does not harm pollinators. But until that happens, you’ll just have to put yourself on the Prairie Moon waitlist and hope.

And without further ado, here’s everything I know about growing downy gentian from seed.

Seed Storage Potential

I haven’t yet tested downy gentian seed storage potential, so I do not know how long the seeds remain viable when kept in warm–dry storage, but I know that they retain substantial viability for at least three months of warm–dry storage.4 I’ve written previously that a different species of Gentiana—bottle gentian (Gentiana andrewsii)—may have a relatively short window of viability.5 Since storage potential for downy gentian is unknown, I would suggest beginning cold–moist stratification as soon as possible (or within three months).

As mentioned in the introduction, moths have a proclivity for downy gentian, boring small holes in the seed pods, eating the seeds, and leaving frass in their place. So I have two pieces of advice for folks harvesting and cleaning downy gentian seeds:

1. Process the seeds the same day you harvest! You may find that heads containing viable seeds also contain larvae. So processing your seeds immediately means you can separate the larvae from the seeds before the larvae eat all of the seeds.

2. Make sure you’re processing gentian seeds, not frass. Gentian seeds are very small, and if you’ve not worked much with them, you may struggle to distinguish them from frass. It would be a shame to waste time cleaning, stratifying, and sowing frass, so grab a loupe and inspect carefully.

Pretreatment and Germination

Stratification

I’ve had such good success germinating downy gentian that I’ve never run an experiment to determine their stratification requirements. In my experience, fourteen weeks of cold–moist stratification yields very high germination. But it’s possible they’ll germinate just as well with a shorter pretreatment time. If I can get my hands on enough seed, perhaps next year I’ll systematically test their germination requirements.

Light Requirements for Germination

I have not tested whether downy gentian requires light for germination. I have, however, tested bottle gentian, which does require light to germinate. I therefore operate under the assumption that downy gentian also requires light. I use a photoperiod of fourteen hours (i.e., in my propagation chamber, the lights are on for fourteen hours per day).

Sowing Protocol

I surface sow my downy gentian seeds, and just barely sprinkle them with vermiculite (truly the lightest dusting you could imagine). The vermiculite is almost certainly unnecessary, but it makes me feel good—it’s like the vermouth in a dry martini: somehow essential, despite being barely present.

I sow mine in SureRoots DeepPots 50-cell trays in unamended ProMix BX potting medium. I do recommend using a deep container (e.g., five inches deep), since downy gentians form a tap root (see Figure 2). With that being said, they will do just fine in slightly shallower containers; they’ll simply form a shorter tap root.

Germination Conditions

As with all my seed germination, I use a humidity dome to keep humidity in the range of 95–100% relative humidity until germination. And as I do with most of my native plants, I germinate my downy gentians using an alternating temperature regime consisting of 20°C days and 10°C nights (more info on alternating temperature regimes here). At that temperature regime, germination typically begins in five to eleven days (though germinants are so small that they are hard to see with the naked eye).

Post-Germination Cultural Practices

Post-Germination Temperature Regimes

After germination, I grow my plants out at the same 20/10°C temperature regime I use for germination, though I allow temperatures to drift as high as 25°C during the day and 15°C at night.

Post-Germination Humidity

I keep my propagation chamber at 76±4% relative humidity, so downy gentian seems to tolerate humidity up to 80%. Alex from Windflower Natives posted about his plants suffering from an excess of humidity, but I don’t have any personal experience of that. Perhaps the humidity in their greenhouses goes above 80%? If you want to be cautious, keep your humidity below 75%!

Post-Germination Watering

I do try to water mine sparingly, allowing the surface of the soil to dry considerably between waterings. Not surprisingly for a plant native to dry upland prairies, downy gentian seems to thrive on neglect. I’ve forgotten to water my seedlings for weeks and saw no seedling death (though I don’t necessarily recommend doing that as a general habit).

Fertility

As with all my plants, I fertilize weakly every time I water. I use 3 ml of Superthrive Grow per gallon of water. I’m generally an organic grower, but I use non-organic fertilizer for my containerized seedlings.

Growth Pattern

I’ve seen reports that downy gentian grows extremely slowly. In my experience so far, relative to their mature size, downy gentian grows at a rate comparable to bottle gentian and cream gentian. That is, after a season of growth, downy gentian rosettes are noticeably smaller than cream and bottle gentians after a comparable season of growth, but since mature downy gentians are also noticeably smaller than mature bottle and cream gentians, I’d estimate that all three species make the same amount of progress relative to their mature size.

Like other true gentians, downy gentian produces only a basal rosette in its first season of growth, and even when environmental conditions are kept constant, that basal rosette will eventually stop growing. I wrote a post recently about this growth pattern, where seedlings behave as though an internal clock stops their growth after a fixed period of time. For lack of a better term, I call this “Type 2” growth, in contrast with the more common “Type 1” growth pattern where species grow indefinitely under favorable conditions. Read the post for more info!

The practical implication of this Type 2 growth pattern is that, no matter how optimal your conditions are—and no matter how happy the plants are—they won’t produce anything more than a few leaves their first season (see Figure 3).

Once the basal rosette has stopped growing, downy gentians need a period of chilling to restart growth. I prefer to do this in my fridge, and I explain my reasoning for this—and the protocol I use—in my post on chilling live plants. After their period of chilling, they resprout with an entirely new morphology; instead of reappearing as basal rosettes, they come out of dormancy as erect stems reaching for the sky, with one to three stems per plant (see Figure 4). I call this transformation “cryomorphosis,” and discuss it in more detail here.

I have no evidence to back up this claim, but I don’t think downy gentians are large enough to plant out after only a single season of growth. The main issue is that, after only one season, they don’t produce enough root mass to hold together the potting medium in a 2.5 × 2.5 × 3.5-inch nursery pot or a 5-inch deep plug. So if you did try to transplant them, the root ball would likely fall apart, and you’d probably damage a lot of fine feeder roots in the process. I prefer to transplant just as the roots are filling out the plug, which doesn’t seem to happen in the first year of growth.

The other issue with planting them out after only one season is that they are not able to compete with other plants. I have observed them to be sensitive to shading, and a basal rosette the size of a quarter can be shaded out by even a tiny weed.

So if you apply my refrigerator chilling method, you can collect seed in October, stratify seeds for three months, sow seeds in January, chill in June, return to spring conditions in September, then chill again in February. Thus by April/May (approximately a year and a half after seed harvest), you have relatively large, third-season seedlings that are ready to plant out (and might even flower that same year).

Planting Considerations

When I’ve encountered downy gentian in prairie remnants, it’s usually growing mixed in with taller plants. Because it has such floppy stems, having stiffer plants to lean on helps keep it upright, and it’s possible it benefits from a little bit of light herbaceous shade. With that being said, it doesn’t like to be fully shaded. I saw one quickly die in my plot when nearby forbs choked out the light. So I would suggest planting it in the company of mild-mannered graminoids, such as little bluestem (Schizachyrium scoparium) or prairie dropseed (Sporobolus heterolepis). If you know of better companions, let me know in the comments!

Well that’s everything I know about growing downy gentian from seed. I’ll be sure to post pictures when my three-year-old plants emerge this spring! ■

Have a question? If you have questions about Midwestern native plant propagation that you want answered in a future post, I’d love to hear from you! Shoot me an email at plantpropagationproject@gmail.com

Want to support this work? Buy me a cup of coffee (or a bag of potting medium)!

Buy me a coffee!

In my previous post, I explained that—although many herbaceous perennials grow more or less continuously under favorable conditions, reaching or approaching mature size in a single season—others allocate only a limited amount of energy to above-ground growth each season, stopping well short of mature size even when conditions remain ideal (see Figure 1). For lack of a better term, I refer to that latter pattern of limited growth, stopping well short of mature size, as “Type 2.”

When producing plugs of native plants, it’s straightforward to handle species that grow continuously until mature size is reached (plants in my Type 1 category): simply grow them out until you’re happy with their size and root mass. Plants with Type 2 growth, however, pose a bit more of a challenge to nurseries and home growers. I like to transplant my perennials when their root mass is just substantial enough to hold together the potting medium of their containers. So what do I do if my plants have stopped growing, but don’t have enough root mass to transplant? Or what if my plants have stopped growing, but are too small to survive competition in the garden/restoration they are destined for?

The simplest option—and likely the most common—is to overwinter these plants, relying on the fact that winter conditions will spur growth in the spring. This process can be repeated for as many years as it takes to produce plants of usable size. This is a perfectly fine method, but I don’t often use it.

Usually, I prefer to chill my plants in a refrigerator because:

• It’s a more efficient use of time. Spring ephemerals, for instance, only actively grow for a couple months, and they spend the remaining nine to ten months of the year dormant. But they often only require two to four months of chilling to break dormancy. So if you’re in a hurry to grow a plant (which maybe you aren’t), why wait nine to ten months for something you can do in two to four months? Artificial chilling, admittedly, takes up energy. But if you have a chest freezer in a cold basement rigged up to an external thermostat to run at refrigerator temperatures (just like a DIY kegerator), you’re likely using a very minimal amount of electricity.

• It’s a more efficient use of space. I have limited propagation space, so space is valuable to me. The more quickly I can get plants matured, the sooner I can repurpose space for a different set of plants. If a tray of plants takes up approximately two square feet of space for three months, then gets transplanted outside, that tray of plants has cost me 0.5 square-foot-years of space. If a tray of plants takes up two square feet of space for three years, it has cost me six square-foot-years. Maybe I’ve lost you on my math, but the point is that I like to minimize the residence time of my plants to keep space open for growing other plants.

• It gives me flexibility in when I have my Type 2 plants in the grow area. My propagation area usually fills up the closer I get to spring planting. So in March and April, plants are squeezed in like sardines, and there isn’t an inch of surface area to spare. In the summer, however, my propagation area is mostly empty. For this reason, it’s very convenient for me to have my Type 2 plants in cold storage from April through May when space is scarce, and then bring them out to grow in June.

• It’s low risk. Plants in the ground are well-insulated by the soil and by organic debris above the soil. So when temperatures drop to ‑30°F (-34°C) here in Wisconsin, the plants in the ground may be only a few degrees below freezing (especially if there is snow). But with containerized plants held above ground, there is much greater risk that the plants will experience extreme cold. It’s true that many native plants can tolerate fairly extreme cold, but I don’t like to take my chances on valuable species.

• I have jumping worms in my area, so I keep my plants mostly indoors. I am perhaps excessively paranoid about jumping worms. But I don’t want to be the guy to spread them further than they’ve already been spread. So my plants only go outside for a brief hardening-off period before transplanting, and even then they are kept well above the ground. So keeping my plants outside for months at a time makes me a bit nervous.

The use of controlled cold exposure to break dormancy and manipulate plant growth is well-established in commercial horticulture. The Dutch cut tulip industry is perhaps the most dramatic example: millions of bulbs pass through climate-controlled chilling facilities each year, enabling growers to force flower production in months in which tulips would not ordinarily bloom. When prolonged cold exposure confers or accelerates flowering competence, the treatment is called—confusingly—vernalization.1 This contrasts with the chilling I describe, which does not confer or accelerate flowering, but releases plants from dormancy so that vegetative growth can be resumed (and—in the case of some lilioids and gentians—induce cryomorphosis). So the chilling I describe is not vernalization in the strict sense, but the technique of chilling is the same technique I would use for vernalization.

Although the principle of controlled chilling has been applied with great sophistication in the ornamental bulb trade, it remains largely unexploited in native plant production, where growers typically rely on natural overwintering and accept a single annual growth cycle as a given (or else don’t grow containerized Type 2 plants at all). That represents a missed opportunity. Many native perennials—especially spring ephemerals, lilioids, and true gentians—have chilling requirements that, once satisfied under controlled conditions, allow plants to be brought out of dormancy weeks or months ahead of natural schedule. For instance, whereas spring ephemerals naturally have a single growth spurt per year, with chilling, they can often undergo two periods of growth in a year, cutting the time to produce mature plants in half.

A Simple Protocol for Chilling Containerized Plants

Chilling plants is sort of like boiling pasta: it might seem exceptionally easy to those with experience, but remember that anything can be daunting if you’ve never done it before. Like pasta, it’s mostly a matter of timing, temperature, and moisture. Below are the steps I take to chill my containerized plants. Some of this info might seem very obvious, but in the interest of making this protocol as accessible as possible, I’m stating even the most obvious steps.

Step 1: Determine that the plant is ready for chilling

If you’re working with a spring ephemeral, this will be fairly easy: just wait until the above-ground growth has senesced (see Figure 2). Once foliage begins to senesce, I generally water sparingly. I’m not sure if it’s really necessary to water at all at this point, but I have an aversion to bone-dry potting medium, so I keep my plants just barely damp. If you’re growing in a controlled environment, be aware that your plants may not senesce as thoroughly as plants outside would. Whereas heat, sun, and wind conspire to utterly desiccate the senesced foliage of spring ephemerals growing outside, no such elements are at play indoors, so leaves may remain green or yellow rather than shriveling up. Generally speaking, this is fine, and there’s no need to wait until full shrivel has been achieved. If you’re working with a plant that is not a spring ephemeral, but does show Type 2 growth (such as true gentians), then you may never achieve full senescence. I typically let these plants grow for five months or so, and then arbitrarily decide it’s time to chill.

Let soil reach just-barely damp

Before chilling, I like my plants to be in potting medium that is just barely damp. At this level of moisture, the potting medium is moist, but no excess water can be squeezed out of it. I have no real evidence that this is preferable, but generally speaking, more air pockets and less water equate to lower mold risk, so I think it’s a good idea.

Seal the containers

Once my potting medium is at the right level of just-barely-damp, I seal up my containers. If I have a whole 1020 tray of containers, I usually just wrap the whole thing in cling wrap. If I have fewer, I put them in a plastic container with a lid. Plastic bags work too. In general, I don’t worry about the lack of airflow. Don’t overthink this part (see Figure 3).

Chill in the refrigerator: ~40°F (~4°C)

Take your sealed pots and put them in the fridge. Do not use a freezer. If you get really into this process, you may want to buy an auxiliary fridge. Domestic partners may not be enthusiastic about plants taking up valuable fridge space. I have not conducted tests on how many days of chilling are required for various species, but I’ve put together some info in Table 1 that combines published chilling requirements with my personal observations.2 I typically default to twelve weeks, though I’m sure I could get away with less for some species. I sometimes open up my containers every four weeks or so to make sure they’re neither too wet nor too dry. This is also a good time to see if there is any moldy organic material to remove.

Remove from fridge

All good things must come to an end, even chilling. After twelve weeks (or however long you’ve determined is appropriate), remove the plants, water them well, and move them to your growing area. Growth usually becomes visible within a couple of weeks. Some plants will begin growing in the fridge when they’ve reached their chilling requirement (e.g., Camassia scilloides); generally speaking, it’s good to get these plants out of the fridge before they’ve grown too much, since in the absence of light, fridge growth can be very spindly and prone to flopping (see Figure 4).

Repeat as necessary

Some plants will only need one cycle of chilling (two cycles of growth) in order to be large enough to plant out. Others will need many cycles. So just keep repeating the grow-chill cycle until your plants are large enough to fend for themselves outside in the garden or restoration. ■

Have a question? If you have questions about Midwestern native plant propagation that you want answered in a future post, I’d love to hear from you! Shoot me an email at plantpropagationproject@gmail.com

Want to support this work? Buy me a cup of coffee (or a bag of potting medium)!

Buy me a coffee!

For the sake of generalization, let’s say that, after seed germination, first-season growth of perennial herbaceous plants follows two basic patterns when plants are given ideal growing conditions (see Figure 1):

• Type 1: continuous growth until mature size is reached

• Type 2: limited growth, stopping well short of mature size1

In the sections that follow, I’ll describe these patterns in further detail.

Two Growth Patterns of Perennial Herbaceous Plants

Type 1: Continuous Growth until Mature Size Is Reached

This is the growth pattern most folks are more familiar with, and it’s typical of plants that are considered “easier-to-grow” (see Table 1). Given a long enough growing season—and optimal conditions—plants in this group are capable of reaching their mature size, or near to it, in a single season of growth. Take, for example, whorled milkweed (Asclepias verticillata). If I sowed seeds today, they would grow continuously, and in a few months, I’d have plants roughly the size of mature specimens (see Figure 2).

Type 2: Limited Growth, Stopping Well Short of Mature Size in First Season

In this growth pattern, plants are frugal with how much energy they’re willing to put into their above-ground growth. These plants are therefore much slower to reach maturity, and they provide a bit more challenge to the grower. For this reason, plants that follow this growth pattern are generally less common in the nursery trade (see Table 2). After germinating, they will devote a certain amount of energy to shoot/leaf growth, and then stop, even when environmental conditions are kept constant (i.e., environmental conditions do not trigger dormancy). So in their first season, they may attain only 5%–25% of their mature size. After they’ve completed their allotted growth for the season, they may visibly senesce (as is the case for spring ephemerals; see Figure 3), or they may simply stop growing above-ground but enter no obvious dormancy (as is the case for true gentians; see Figure 4).

For Midwestern perennials, after growth has ceased for the season, additional growth can typically be triggered by a period of chilling. This can be provided naturally through overwintering, or it can be simulated by the grower in a refrigerator. But that’s a topic for a future post.

Introducing Cryomorphosis

For those plants that stop growing after a certain amount of growth, a subset of those plants emerges from their first winter—or simulated winter—looking like entirely different species. The Michigan lily (Lilium michiganense), for instance, spends its first season as one or two modest leaves, but after a period of chilling it presents itself as a tall, whorled-leaf stalk (see Figure 5). Similarly, downy gentian (Gentiana puberulenta) spends its first season as a basal rosette, but after a period of chilling, reemerges as one or more erect stems (see Figure 6).

I think this phenomenon deserves a name, so I’m calling it cryomorphosis. I’ve chosen “cryo-” as the prefix here because chilling is the mechanism I have observed, but note that the underlying trigger may prove to be dormancy release rather than cold exposure per se. Also note that I’m naming an observable phenomenon, not some underlying biological reality. And this phenomenon is not a true binary; the degree to which a plant’s morphology is changed by dormancy is variable. I’m not claiming that this concept is particularly useful for botanists, or that it refers to a single mechanism, but as a grower, it’s useful to have a vocabulary to describe the patterns I observe in plant growth.

It’s common for perennials to have juvenile forms that look radically different from their mature forms. In botanical terms, the shift from a juvenile form to mature form is called heteroblasty. Heteroblasty refers to any shift in morphology, including the gradual shifts that occur across a season. But since cryomorphosis is an apparently sudden phenomenon, seemingly tied to dormancy, I think it deserves its own special name.

Understanding this facet of the life cycle of native plants can help us strategize our propagation approach and set expectations. When I started growing lilies and gentians, I thought it was because of my lack of skill that my plants were stuck in their juvenile form, refusing to send up a stalk. Now that I understand their growth strategy more, I’m able to plan my propagation schedule around cycles of growth and dormancy. Moreover, I find that the juvenile form of plants can be very vulnerable in the garden setting, with even the slightest amount of shading/crowding capable of killing a small basal rosette. So I aim to plant my lilies and gentians after they have formed an erect stalk, and I know that doing so requires subjecting them to a chilling cycle.

Cryomorphosis Is Not Vernalization

One last thing: for those readers who are thinking to themselves “isn’t this just vernalization?” my answer is that no, cryomorphosis is distinct from vernalization. Vernalization refers specifically to the cold-induced acquisition of flowering competence.2 That is, the process by which a plant becomes capable of flowering after a period of chilling. This is why winter wheat needs to be sown in the fall and experience cold temperatures in order to flower and produce grain, why artichokes generally won’t flower until after a period of chilling, and why daffodils flower poorly in mild climates unless they are pre-chilled before planting. So while vernalization is about flowering, cryomorphosis is about plant architecture. A gentian that sends up a central stalk after chilling has not necessarily acquired flowering competence; it has simply transitioned from its juvenile basal form to its mature caulescent form. Likewise, a Michigan lily sending up its whorled-leaf stem after its first winter has undergone cryomorphosis, but it may not flower for another season or two. ■

Have a question? If you have questions about Midwestern native plant propagation that you want answered in a future post, I’d love to hear from you! Shoot me an email at plantpropagationproject@gmail.com

Want to support this work? Buy me a cup of coffee (or a bag of potting medium)!

Buy me a coffee!

Two factors in particular make this plant a little bit more challenging to successfully grow:

• Its seeds appear to be desiccation-sensitive. I’ve discussed this trait previously in my post on growing bottle gentian (Gentiana andrewsii) from seed, so check out that article if you want to learn more about desiccation-sensitivity. Pedicularis canadensis seeds appear to rapidly lose viability when stored dry at room temperature.2 So if you’ve bought Pedicularis canadensis seeds, stratified them, and watched them mold, the problem was probably the seed, not you.

• As a hemiparasite, it requires a host plant to thrive. I’ve discussed this trait before in my post on Castilleja coccinea. Pedicularis canadensis is likely an obligate hemiparasite, meaning that—despite having the ability to photosynthesize—it requires a host plant to complete its life cycle.3 I’ve seen a lot of people say that it prefers graminoids (grasses, sedges, rushes) as hosts, but I don’t really believe that. We’ll leave that topic for a future blog, though.

So to successfully grow P. canadensis, three conditions must be met:

1. The seed must be fresh. Satisfying this condition is simple: to grow Pedicularis canadensis from seed, you’ll either need to find someone who will sell you fresh seed, or collect the seeds yourself (please follow all local regulations concerning gathering plant material). How fresh is fresh enough? I don’t know. With four-month-old seeds, I got 0% germination. This Grassland Restoration Network post says they got good germination for seed stored one month. So viability seems to remain high for the first month, but drops off somewhere before month four. This coming year I’ll test how viability changes over time, so stay tuned!

2. The seed must be subjected to an appropriate stratification regime. What is the appropriate stratification regime for Pedicularis canadensis? Read the rest of this article to learn about my results this past year.

3. The seedlings must be grown out with a suitable host species. What is the ideal host plant for Pedicularis canadensis? I’m not sure, but I’m currently growing my Pedicularis canadensis seedlings with these eleven prairie plants: Artemisia campestris, Bouteloua gracilis, Dalea purpurea, Geum triflorum, Ionactis linariifolia, Liatris cylindracea, Schizachyrium scoparium, Solidago ptarmicoides, Sporobolus heterolepis, Symphyotrichum oolentangiense, and Viola pedata. In a few months I’ll take measurements and post the results. According to Piehl (1963), Pedicularis canadensis is not very picky about what plants it parasitizes,4 but in my experience growing Orobanchaceae in a greenhouse setting, choice of host plant can make a difference, even when the hemiparasite is a generalist.

Two stratification regimes for germinating Pedicularis canadensis

Assuming that you have procured fresh-enough Pedicularis canadensis seeds, the next step in growing them will be to subject them to stratification. I heard from Alex Blondeau at Windflower Natives that P. canadensis responded well to both cold-moist stratification (~six months in the fridge) and to warm-moist followed by cold–moist stratification (~three months at simulated summer temperatures followed by three months in the fridge).

So I decided to test both of these methods on fresh seeds from the same lot to see if one method would have better germination.5 All seed was collected June 21, 2025 (this was late in the season, more on that later). Ten days after collecting my seeds, I began pretreatment:

• 219 seeds received six months cold–moist stratification. The seeds were mixed with damp play sand and stored in the fridge for six months.

• 219 seeds received three months warm–moist stratification followed by three months cold–moist stratification. Seeds were mixed with damp play sand, then subjected to simulated summer temperatures (in the dark) for three months (30°C days, 15°C nights). Then they were moved to my refrigerator for three months.

Both groups were treated for a total of six months. After their treatment period, I sowed them on ProMix BX, misted them with water to wash away the sand and settle the seed, and dusted them with fine vermiculite. They were left at simulated spring temperatures (20°C days, 10°C nights) with 14 hours of light and 10 hours of darkness per day.

At sowing time, many of the cold-only seeds had germinated in the fridge, but none of the warm-then-cold seeds had germinated. After sowing, the cold-only seeds began emerging after only a few days, but the warm-then-cold seeds took two and a half weeks to start emerging. I evaluated germination at two points in time: first, at three weeks after sowing; second, at five weeks after sowing. I only calculated statistical significance for the percentages at five weeks. Below are the results:

Ultimately, there was not even a whiff of statistically significant difference between the two treatments when measuring germination five weeks after sowing. However, the cold-only group was quicker to germinate, with germination complete at week three, whereas the warm-then-cold group was just starting to germinate at week three (but note that timing differences were not analyzed statistically). Both groups only achieved about 10% germination, which is much lower than I usually target. I can’t say with certainty why that happened, but here are my thoughts:

• I collected my seeds late—about 2 weeks later than usual. Most of the seeds had already fallen out of their capsules, so I was collecting the small fraction of seeds that remained. It’s possible these stragglers were either less viable or more deeply dormant than the general population of seeds.

• Given that many of the cold-only seeds had germinated in the fridge, I believe six months is too long a period for cold–moist stratification. By the time I sowed seeds, some of the germinants may have already died from spending too much time in the fridge.

• Conversely, given that it took the warm-then-cold group two and a half weeks to start germinating, dormancy was not totally broken at the time of sowing. So they likely needed longer stratification (my intuition is they needed more time in the fridge).

• I essentially surface-sowed the seeds. It’s possible that light inhibits germination (see my article on passionflower for more info on negative photoblasticity).

• I only tested germination at simulated spring temperatures. It’s possible they prefer simulated summer temperatures to germinate.

• I only stratified them at simulated summer temperatures. It’s possible they need stratification at simulated summer temperatures, then simulated fall temperatures, or else just simulated fall temperatures.

• I stratified seeds ten days after collection. It’s possible they require some period of warm-dry after-ripening (e.g., one month).

Rest assured I will test each and every one of these possibilities in the coming year! But if you’ve got a few hundred seeds, and only need a few dozen plants, either of these protocols could work for you.

But isn’t it odd that Pedicularis canadensis seeds would respond to both cold and warm-then-cold stratification?

It might seem odd to you that a species would respond to both cold and warm-then-cold stratification. But it doesn’t seem very odd to me, and there are at least two possible explanations for this phenomenon:

• Hypothesis 1: Pedicularis canadensis seeds have intermediate physiological dormancy. Pedicularis canadensis seeds are described by Baskin and Baskin (2014) as being physiologically dormant,6 which means that they have a physiological inhibition mechanism in the embryo; this inhibition often manifests as insufficient embryo push power to overcome covering-layer restraint.7 Baskin and Baskin describe three levels of physiological dormancy: non-deep, intermediate, and deep. Many seeds with intermediate physiological dormancy have the interesting trait that their relatively long cold stratification requirements can be reduced by a period of warm-moist stratification. One explanation for this trait is that it allows species adapted to long winters to extend their range to areas with relatively short winters. The precise level of dormancy that Pedicularis canadensis seeds have has not yet been described in the academic literature, but it’s possible that P. canadensis has intermediate physiological dormancy.

• Hypothesis 2: the warm–moist stratification isn’t breaking seed dormancy—it’s simply keeping the seeds moist until cold stratification begins. Put another way, the group receiving warm-then-cold stratification was essentially held in moist storage for three months, then cold–moist stratified for three months. This could explain why the warm-then-cold group was so much slower to germinate: the seeds received less time experiencing conditions necessary for breaking dormancy (near-freezing).

If all goes according to plan, in about one year I’ll have experimental results indicating which—if either—of those hypotheses is correct. ■

Have a question? If you have questions about Midwestern native plant propagation that you want answered in a future post, I’d love to hear from you! Shoot me an email at plantpropagationproject@gmail.com

Want to support this work? Buy me a cup of coffee (or a bag of potting medium)!

Buy me a coffee!

It’s January, which means it’s time to start sowing native plant species so they’ll be large enough to plant out in spring. Some of my slowest species I sowed in October and November (e.g., bottle gentian, American pasqueflower), and I won’t sow fast-growing—and heat-loving—plants until February or March. But January is a good time to sow plants that grow at a moderate speed; prairie smoke, wild geranium, and native violas, for instance, take about 14 to 16 weeks to fill out a standard 2.5 × 2.5 × 3.5 in. nursery pot.

I get a lot of questions about the setup I use for growing native plants, so I figured I’d cobble together a quick post to address the most important aspects of an indoor plant setup. I won’t describe here the exact setup I use (it’s overly complex for most folks), but I’ll describe a setup that I used when I started out growing plants indoors (see image below), which I think will work for most people, and is cheap and easy to set up.

To grow native plants, you only need three pieces of equipment: lights, pots, and trays. I’ll describe below what to look for in each, and I’ll conclude with some nice-to-haves that can improve your experience.

Lights for Growing Native Plants

Help! Which lights should I buy for my plants?

The internet is awash with misinformation about grow lights. Especially with the rise in home-grown cannabis, there are hundreds—maybe thousands—of grow light vendors competing to convince you that you need their products. But you don’t need their products, and your plants are unlikely to benefit from them.

For the average home grower raising seedlings, the best lights are usually the cheapest ones you can get your hands on. For years, I used the cheapest LED workshop lights available at my nearest big box hardware store (Menards). As of this writing, Menards sells the Pinegreen Lighting 46” 6500 Lumen LED Shop Light for $7.99. That’s the type of price you should be paying. The exact light doesn’t matter. Usually the cheapest lights are approximately four feet long. If you don’t have space for four-foot-long lights, you can usually find two-foot-long workshop lights (but they’re often a little pricier).

How many lights do I need for my growing setup?

Generally speaking, I suggest using three four-foot-long workshop lights to cover a two-foot by four-foot growing area (eight square feet). So if you plan to grow four 1020 trays of plants (1020 trays are the standard 10 × 20 in. nursery trays), you need three workshop lights.

How many hours of light do my plants need per day?

Different plants have different light requirements. Some plants, such as members of the milkweed (Asclepias) genus, can go dormant when day lengths are too short.1 But plants value their sleep, so too long of a day can also cause problems.2 I use a photoperiod of 14 hours on, 10 hours off for all my plants, and I think that’s a good middle ground that works for most species. You could turn your lights on every morning when you wake up at 6 a.m., and turn them off before bed at 8 p.m., but I suggest purchasing an outlet timer. Again, just buy the cheapest one. Just make sure that if your lights have three-prong plugs, you buy a timer with a three-prong socket.

But what about lumens?

Lights are marketed in terms of their lumen output. Lumens are a measure of how bright a light appears to the human eye, but plants are not human and they don’t have eyes. Lumens measure the intensity of light weighted by the human photopic sensitivity curve, which peaks around 555 nanometers. This wavelength was chosen because our eyes are most sensitive to green light, meaning that green light appears brighter to us than other wavelengths. Plants are green because—to oversimplify—they reflect more green light than they absorb; chlorophyll absorbs red and blue light more strongly. Because lumens are weighted to human vision (peaking in the green), they emphasize wavelengths that often contribute less per photon to photosynthesis than red light (see diagram below).3

If you start looking at fancy grow lights that have more output in the red/far-red spectrum, those lights might have very low lumen values (violet and far-red light appears dim to humans), even though they are optimized for plant growth. And in many cases, professional grow lights will often not list lumens, since vendors want to avoid an apples-to-oranges comparison to human-centric lighting.

Money is no object to me, so just tell me, what are the nicest lights I can buy?

I’m not really a lighting expert, so you shouldn’t trust me. Cheap lights are inefficient, both in terms of how much energy they waste as heat, and in terms of how useful their wavelengths are to plants. The Menards lights I mentioned above are fine, but maybe they get a little hotter (wasting energy) than nicer lights, and they may be outputting too much green light and not enough red and blue light. I’ve been using lights from ActiveGrow, and they’ve worked well. I haven’t necessarily noticed improvements in my plants, but they have some nice design features that make them easy to work with, and ActiveGrow provides graphs of the relative intensity of wavelengths outputted (though I don’t know if those are third-party verified).

Trays for Growing Native Plants

This part is easy. Just get yourself some 1020 trays, without holes. 1020 (pronounced “ten twenty”) trays are, as the name implies, 10 × 20 in., and they are a standard size. They cost about $1.00–$4.00 each depending on where you buy them (your local hardware store or garden store probably has them). If you want to upgrade, buy the rigid type that will never bend or break. I use “Indestructible Trays” from Neversink Tools. Their customer service is, at times, a little salty, but they make the best tools so they can be forgiven.

Pots for Growing Native Plants

In general, deeper pots are better. I plant many of my natives in places where I do not have access to water. So I plant plugs that are five inches deep, meaning that root tips are five inches below the soil surface. The soil five inches deep is usually moist throughout the spring (where I live), so I hardly water my plants. If your pots are only three inches deep, you’ll probably need to water them more, because the soil is likely to dry out at three inches.

With that being said, if you’re on a budget, just make friends with gardeners and collect their used pots. The pots I prefer to use are the 2.5 × 2.5 × 3.5 in. size that you see most often at native plant sales. If you can’t find friends with used pots, Greenhouse Megastore sells packs of 32 for $2.88. Or just use another type of pot. The advantage of this type is that they are big enough to accommodate most native plants until they’re ready to be planted, but they’re small enough that you can fit 128 of them in a 2 × 4 ft. growing area.

My favorite container for native plants is the SureRoots Deep 50 Cell Plug Tray. Each plug has approximately the same soil volume as the 2.5 × 2.5 in. pots mentioned above, but they are deeper and narrower. Deeper means that your roots will be deeper in the ground at planting, which means less need to water. And narrower means you can fit more in a smaller space. So you can fit 200 plants in a 2 × 4 ft. growing area.

Bonus Items: Shelving and Fans

You don’t need shelving and fans, but they’re nice to have. With a nice shelving setup, you can grow three or four shelves of plants right on top of another, and it gives you an easy way to hang your lights. Wire shelving racks work well, and you can often find them cheap.

I like to have a very low-power fan going on my plants throughout the daytime. Constant airflow keeps fungus gnats at bay, reduces risk of fungal infection, and it makes for more resilient plants that can transition more easily to the outdoors. I favor low-power computer fans, like this one from AC Infinity. Just keep them at the lowest setting. Make sure to be careful about acclimating freshly-germinated seedlings to the fan. I usually wait until seedlings are one to two weeks old before exposing them to the fan, and I usually start them off in an area farthest away from the fan, where airflow is least intense.

Have questions? Let me know in the comments!

Leave a comment

To answer that question, we first need to establish what prairie we’re talking about. On the one hand, it is possible to discern a discrete 200-square-mile grassland from the 1830s land survey. But while that shape exists on paper, the decision to consider just that single treeless expanse as the prairie is arbitrary. In the 1830s, a belt of prairie stretched from Madison to Lake Winnebago (see figure below).1 Although 1830s surveys show that there were interruptions to the prairie, these interruptions were largely oak savanna (and sometimes very thinly timbered oak savanna at that). So when we read descriptions of the prairie north of present-day Madison, we can’t be sure if writers were describing the 200-square-mile chunk of uninterrupted grassland that was captured on early survey maps, or the belt of prairie running to Lake Winnebago.

As for the original name of the prairie (whatever its boundaries may be), one of the earliest names I’ve been able to track down comes from Juliette Augusta Kinzie’s 1856 memoir, in which she chronicles her 1831 return trip to Fort Winnebago from Chicago. That journey included a trip across the prairie belt (in the area now considered the Empire Prairie), which she says was known as the “Twenty-mile Prairie.”2 While the name “Twenty-mile Prairie” might suggest to modern readers that the prairie extended twenty miles in every direction, I’d argue it’s equally plausible that it was named simply for the fact that it took twenty miles to cross traveling north–south, with no judgment on its northeast–southwest length.

The earliest recorded name for the prairie that I’ve found comes from Starin’s 1840 diary of his travels in Wisconsin. In addition to writing about the poor moral character of Madison’s residents (“The people are very avaricious, are professed enemies & take every possible advantage of strangers”), he also describes a twenty-three-mile journey across what he calls “Winnebago prairie.”3 “Winnebago” is an exonym for the Ho-Chunk people, and was commonly used in the 1800s by White settlers. So translated to contemporary language standards, Starin was really calling the grasslands north of Lake Mendota “Ho-Chunk Prairie.” It’s not clear to me if that was a commonly-used name for the prairie, but it’s not surprising that the grassland would come to be associated with the Ho-Chunk. As shown in the map below, at the time of White settlement in Wisconsin, the Empire Prairie—and the rest of the prairie belt it belonged to—was within Ho-Chunk lands.

By 1832, the U.S. government had coerced the Ho-Chunk into ceding all of their territory east of the Wisconsin River (including the prairie belt mentioned above), and in 1837 the U.S. government coerced the Ho-Chunk to cede all remaining territory within present-day Wisconsin (see map below). Despite the theft of their lands by the U.S. government, many Ho-Chunk remained in—or continued visiting—lands east of the Wisconsin River. It’s not clear whether the grassland Starin called “Winnebago Prairie” was named in recognition of its location on Ho-Chunk soil, or because of continuing Ho-Chunk presence in its vicinity. In his 1877 history of Burke Township, John Douglas, writing about a trail through the prairie, claims that “early settlers frequently saw as many as five hundred to one thousand Indians, with their wives and little ones, pass and repass in a week.”4 I’m reluctant to take Douglas’s words at face value (late 1800s recollections of American Indian history can be a bit imaginative), and Douglas provides no insight into which Indigenous nation the travelers belonged to. But taken with Starin’s naming of the “Winnebago Prairie,” it suggests a Ho-Chunk presence in the grasslands north of Madison.

Between the Four Lakes and the Portage

I have not yet found any Ho-Chunk names for the so-called “Empire Prairie,” or for the band of highland prairie that stretched northeast from present-day Madison. But the Empire Prairie lay between two areas with significance to the Ho-Chunk: Teejop (the Four Lakes area) and Wawa’ąįja (the portage between the Fox and Wisconsin rivers). In the early 1830s, Teejop was home to at least four Ho-Chunk villages, and the area west of Wawa’ąįja was home to at least three villages along the Hoguc (Baraboo River).

Beyond Douglas’s claim of Indigenous travelers on the prairie between Four Lakes and Portage, I have found additional indirect evidence that Ho-Chunk people traveled across the so-called “Empire Prairie” with relative frequency. In an 1829 map, a path is drawn between Four Lakes and Portage, which certainly suggests a Ho-Chunk road. It’s possible this was a road created by White settlers to get from Portage to Blue Mounds, but given that the first White settler had arrived in Blue Mounds only the prior year (to illegally occupy and mine unceded Ho-Chunk lands), that seems less likely.

When the Empire Prairie region was surveyed in the early 1830s, a road between Four Lakes and Portage was included on official plat maps and labeled as the road between Fort Winnebago and Blue Mounds. I cannot say with certainty that the Blue Mounds–Fort Winnebago road is the same trail as in the 1829 map, but there is strong evidence: the 1829 map labels a pond ten miles north of Four Lakes, and the Blue Mounds–Fort Winnebago road intersects a pond ten miles north of Four Lakes (and ponds are very rare in the upland prairie). We also know that a road from Blue Mounds to Fort Winnebago existed in 1829, because that road was used to define the western boundary of Ho-Chunk territory in the 1829 treaty in which the Ho-Chunk ceded one-third of their lands to the U.S. government. Amazingly, the path of this historic road runs right through Dane County’s largest remnant of the Empire Prairie: Hauser Road Prairie (see map below).

Parts of the Blue Mounds to Fort Winnebago road would later be incorporated into the Old Military Road, which connected Fort Howard (in Green Bay) and Fort Crawford (in Prairie du Chien) by way of Fort Winnebago. I have seen many claims that the Old Military Road largely followed established Ho-Chunk trails, but I have not been able to verify these claims. H. E. Cole, for instance, in his 1925 article on the Old Military Road, writes that the road largely followed “well-trodden Indian trails,” but Cole also claimed that the road skirted the northeastern shore of Lake Mendota, which Durbin and Durbin show to be false.5 So Cole is not a reliable source. Durbin and Durbin, for their part, make their own claims about roads replacing Ho-Chunk trails, describing an “existing wagon trace running along the Indian trail from Prairie du Chien.” But their supporting evidence mentions only an existing road and makes no mention of its origin (Ho-Chunk or otherwise).6 So while it’s probable that the trail existed before White settlement, I have not found direct evidence of that fact.

James Doty, one of the road’s planners, made a point of contrasting the straight line of the Old Military Road with the existing trails, writing that “travelling has been blindly confined to the old Indian trails, which frequently lead, by very circuitous and unfavorable routes, from point to point, visiting in their course villages and other points out of the general course of the line.”7 Doty, for his part, may have been exaggerating the state of existing roads for the sake of drumming up support for his new road. But if Doty is reliably reporting the existence of Indigenous trails, then perhaps the path from Blue Mounds to Fort Winnebago, with its detour to Four Lakes, is one of the “circuitous” Indigenous routes Doty bemoaned.

The other major recorded route through the prairie—possibly a Ho-Chunk road—is a trail passing Token Creek Springs, first attested to in an 1845 deed of sale as the “Fort Winnebago road.”8 Plat maps from the early 1830s show an unnamed road running from Four Lakes to Token Creek Springs, but do not show it as extending north beyond that point (see image above). I have not found any historical maps clearly labeling that road, but there still exists a road called “Portage” running through Token Creek, so we can hypothesize it follows the old route to Portage. I’m not sure that this route to Portage followed the original trail, given that the trail recorded in the early 1830s from Four Lakes to Token Creek Springs is a bit east of the road currently known as Portage, but it seems plausible that there was a Ho-Chunk road in that general area. John Douglas, writing in 1877, claimed that “the Indian trail from Koshkonong and Third lakes [sic], to Fort Winnebago (now Portage City), passed through the town, near the Indian Spring, on section 26. It was a well beaten track, and is still seen, where cultivation has not obliterated its traces.”9 The road from Third Lake (present-day Waubesa) to Token Creek Springs is recorded in the 1830s survey, so there is corroborating evidence for that leg of the road, and it seems highly plausible that it was a Ho-Chunk travel corridor. The leg from Token Creek Springs to Fort Winnebago is not recorded on early maps, so we’d have to take Douglas at his word (which, as mentioned, I’m reluctant to do).

Douglas goes on to claim that the area around Token Creek Springs, which lies along one edge of the prairie, had once been an “old Indian camping ground.”10 Again, Douglas is not very precise, and it’s not entirely clear what he means by that. He could be saying that the Ho-Chunk had resided there in recent memory, or possibly he means that archaeological evidence (such as mounds) suggested prior habitation. Indeed, he does mention two circular mounds within a mile of the putative camping ground.11 Either way, while it seems entirely plausible that it was a camping ground at some point, we cannot be sure that the Ho-Chunk had resided there in recent centuries.

Anyway, that’s all I have for today. I’ve done my best to research the Ho-Chunk history of the Empire Prairie, but there’s more work to be done. Let me know in the comments if there’s anything I’ve missed!

In the 1830s, southern Wisconsin was dominated by a vast savanna, and within that savanna was a near-treeless expanse of prairie (now called the Empire Prairie) that stretched across 129,000 acres. Today, only 168 acres of the Empire Prairie remain, one-tenth of one percent of what once blanketed the landscape. Those surviving remnants—along with about 37 acres of savanna within the 1830s boundaries of the prairie—support an extraordinary diversity of plant and animal life. Hauser Road Prairie alone is home to at least 148 plant species (including some that are state-endangered or threatened), along with animal residents that are generally rare in southern Wisconsin: meadowlarks (Sturnella magna), dickcissels (Spiza americana), badgers (Taxidea taxus), and a number of leafhoppers that can only be found in prairie remnants.

The phrase “prairie remnant” may mean something slightly different depending on who you ask, so I’ll do my best to briefly sketch its meaning. There are three factors that affect a patch of prairie’s remnant status:

1. Cultivation history (plowed vs. unplowed)

To be a remnant, a patch of prairie must, at the very least, have never been plowed. When farmers started plowing the prairie, they focused their efforts on flat, mesic soils where crops would grow easily. Steep hills with rocky, dry soil were not worth the effort to try to farm, and were therefore saved from the plow. For this reason, many of the Empire Prairie remnants are on steep hills with dry soil, and this can lead to the false impression that the Empire Prairie was, as a rule, very dry (dry or dry-mesic). It’s true that Hugh Jamieson’s report of crossing the Empire Prairie supports the notion that water was scarce, but certainly there existed sections of the prairie that were mesic (i.e., moist but well-drained). But since those mesic areas were plowed, the remnants we have of the Empire Prairie mostly lack species adapted to mesic soils: purple milkweed (Asclepias purpurascens), Sullivant’s milkweed (Asclepias sullivantii), wild onion (Allium sp.), Michigan lily (Lilium michiganense), prairie dock (Silphium terebinthinaceum), and wild quinine (Parthenium integrifolium), to name a few. It’s possible those species never lived on the Empire Prairie, but it also seems likely that they did live here and were locally extirpated when their habitat was converted to agriculture.

2. Prior fire regimes

In the absence of fire, prairies are colonized by shrubs and trees. Some grassland seeds can persist in the soil for many years, and then germinate when the trees are cleared by fire (or mechanically). But most seeds can’t stay viable for decades, so the longer a patch of land has been overtaken by woody plants, the less likely it is to rebound naturally to prairie. In southern Wisconsin, it was common practice to burn pastures to promote tender green growth for animals to graze on. It’s difficult to say exactly how common burning was, and when precisely pasture management moved away from burning, but Hugh Jamieson presents it as common on the Empire Prairie in the 1800s,1 and we know that pastures were burned in Dane County well into the 1900s.2

Although I don’t know the burn history of the Empire Prairie remnants, it seems likely that a number of them were managed with fire until the mid-1900s. If this is true, then when conservation efforts began in the 1970s,3 some remnants may have only experienced a few decades of fire suppression, which could explain the diversity of species in these prairie remnants. Indeed, when looking at historical aerial imagery of these remnants and surrounding areas, it is striking how little woody encroachment there was in 1937 and 1955 (more on this below), which is consistent with the idea that fire was commonly used for the first half of the 20th century.

3. Past grazing intensity

Obviously prairies can withstand some grazing; herds of bison numbering in the tens of thousands previously grazed western grasslands, and smaller herds grazed in what is now Wisconsin, and probably in the Empire Prairie (see map below).4 But that grazing was sporadic, and only a small fraction of the landscape was grazed at any given time. Consider the example of bush clover (Lespedeza sp.), which bison are known to eat. Even if bison removed all the bush clover from an area, neighboring populations could recolonize the depleted patch. This contrasts with intensive pasturing practices that farmers often use, where grazing animals are confined to a relatively small area, and graze it over and over again. In this situation, once the bush clover has been removed, there is no surrounding population to recolonize, since the pasture is surrounded by agricultural fields.

While many prairie remnants have experienced some grazing, some were heavily grazed, and this manifests today in floristic communities with relatively low diversity. When ecologists try to determine past grazing intensity, they often turn to indicator species. Compass plant (Silphium laciniatum), for instance, tolerates light grazing, but will disappear under heavy grazing pressure. So its presence in a remnant suggests a history of relatively light grazing.

With all this in mind, remember that remnants exist on a spectrum. On one end are remnants that have been burned consistently and grazed minimally since White settlement, and on the other end are remnants that have been overtaken by woody plants or grazed heavily (or both). They’re all remnants, but the more degraded remnants may have lower biodiversity. It’s worth noting that many prairie remnants are actually restored to some degree. Beyond regular burning, to help remnants rebound, land stewards mow aggressive weeds, selectively remove woody species, and introduce seeds of missing species (usually from close by).

Mapping the Remnants of the Empire Prairie

I’ve been posting a lot about the Empire Prairie, and perhaps you feel inspired to visit a remnant of this historic prairie. To the best of my ability, I have mapped out the documented remnants of the Empire Prairie.

To determine the boundaries of documented Empire Prairie remnants, I inspected 20th-century aerial imagery5 to determine what areas had been plowed in the previous century. Of course, it’s possible an area was plowed in the 19th century, then allowed to return to grassland in the 20th, so the methodology isn’t perfect. In some cases, like the example below from the Hauser Road Prairie, the current boundaries of the prairie include areas that were at one point plowed, but have reverted back to prairie without human intervention beyond prescribed burning. For the purposes of this cartographic exercise, I consider those areas to be prairie restorations, not remnants.

Below are some more detailed maps of the Empire Prairie’s remnants (as well as remnant savanna). Note that, for the purposes of these maps, I consider prairie with woody encroachment to be savanna, so these maps lack some nuance. I’ve also mapped out prairie restorations that I know of. Not all these areas are publicly accessible, so make sure to do your research before visiting one.

Southwestern Remnants

The oak savanna at Frederick’s Hill lies right at the edge of the 1830 prairie boundary, so if you stand on that hill and look north toward the new prairie restorations at Pheasant Branch, you can try to imagine what the savanna–prairie boundary once felt like. The area just south of the Pheasant Branch prairie restoration is currently wetland, which might be surprising, given that surveyors described that area as prairie. But keep in mind that the water level of Lake Mendota has been artificially raised nearly five feet, and that little pocket of prairie is only a half-mile upstream from the lake’s edge. My suspicion is that area was historically dry prairie, but now drains poorly with Lake Mendota’s higher water level.

Westport–Waunakee Remnants

The Westport Prairie is probably the best place to visit if you’re new to prairie remnants connoisseurship and want to dip your toes in. Unlike many of the Empire remnants, it has an extensive trail system, and the north drumlin has very dramatic topography with expansive views. Plus it is home to many rare plants. The Westport Prairie previously consisted of four disparate remnants, but over the last decade, Groundswell Conservancy has been using seeds collected from within the remnants to expand the prairie into the adjacent croplands. Today, the remnants are mostly linked together, and in the coming years they’ll expand the prairie even further.

North of Waunakee Remnants

In my opinion, the Hauser Road Prairie—owned and managed by The Prairie Enthusiasts—is the crown jewel of the Empire remnants. I’ve flattered it enough at the beginning of this post, so I won’t wax poetic any further. Waunakee is expanding northward, so there is a real risk that the farmland adjacent to the Hauser Road Prairie will eventually be converted to housing subdivisions, preventing any possibility of expanding the prairie through restoration. If you want to donate directly to the Hauser Road Prairie, you can do so by pressing the button below!

Donate to the Hauser Road Prairie

Goose Pond Area

There’s only one remnant here: The University of Wisconsin’s Arlington Farm Prairie. North of that is probably the most ambitious restoration project in the former Empire Prairie, concentrated around Goose Pond (recall that Hugh Jamieson described this as the only source of water in the prairie).

Mud Lake Remnants

The Hagen Unit prairie is the largest remnant of the Empire Prairie, but I cannot vouch for its quality, since I have not yet made the trek out to see it. Note that the original plat maps were not wildly accurate, so even though the Mud Lake Unit prairie appears to be mostly outside the 1830s boundary, that could just be because of error in the map drawing process.

Finding remnants with satellite data

Using the information I have on documented prairie remnants and restorations, I trained a machine learning algorithm to identify other possible remnants of the Empire Prairie using freely-available satellite imagery and topographic data. Based on that work, I’ve identified a few additional remnants, but I won’t map them here since they’re all on private property. But if you’re a conservationist looking for landowners to reach out to, let me know! ■

Have a question? If you have questions about Midwestern native plant propagation that you want answered in a future post, I’d love to hear from you! Shoot me an email at plantpropagationproject@gmail.com

Want to support this work? Buy me a cup of coffee (or a bag of potting medium)!

Buy me a coffee!

Start Survey

First, let me acknowledge that it can be hard buying gifts for gardeners. The items us gardeners yearn for can be hyper-specific: a particular plant, a particular type of seedling cell tray, a particular inline fan to ventilate our germination chamber. I’ve tried to focus on items that appeal to a wider swath of gardeners, but use your best judgment: some of the items here appeal to a relatively small subset of gardeners.

Tools that make good gifts for gardeners:

Niwaki Forged Trowel

This trowel is the best tool I own, and I use it for basically every hole I dig. I’m rough on my trowels (and I have snapped many), but mine shows no signs of wearing down, even after years of abuse. It’s also larger than a typical trowel, so it’s a great tool for jobs that require something between a shovel and trowel (e.g., digging holes for gallon-sized plants)

Niwaki Hand Forged Damascus Hori Hori

I don’t think this needs any explanation. It’s the diamond ring of gardening, an unreasonably expensive metallurgic masterpiece. And make sure to read the backstory on the website.

Hori Horis are wonderful tools for precision work in the garden, but make sure not to pry with them (they will bend or break). If you like the idea of gifting a hori hori, but don’t want to commit to this one, Niwaki offers a less extravagant forged one, as well as a variety of stainless steel options.

Hand Pruners

For gardeners doing a lot of pruning, it’s worth it to have a good pair of hand pruners. Niwaki, again, offers many very aesthetic options.

Alternatively, Felco pruners work great, and my pair has lasted me eight years and shows little sign of slowing down.

Silky Handsaws

If you’re pruning trees and shrubs, you need a good handsaw. Speaking as a (formerly) ISA-certified arborist, I believe Silky makes the best pruning saws. The Tsurugi straight saw works great for detail work, and for more heavy-duty work I’d suggest the Tsurugi curved. Whatever you choose, you can’t go wrong with Silky.

Neversink Interchangeable Hoe

If you need to control weeds across a large area, this is the tool you need. Many different quick-change head options allow you to target different types of weeds growing among different crops.

Neversink Flame Weeder

If your loved one is a pyromaniac, or is the type of person who likes to solve simple problems with extreme measures, they might love a flame weeder. For vegetable gardening, you use this to prep beds, or to weed between rows. For a prairie planting, you could try using this in lieu of controlled burn (I would remove all dry garden debris, then run this a few times in the fall or spring to kill weed seeds at the soil surface).

Strictly Medicinal Seed Screens

If your loved one collects and saves seeds (especially native seeds), this set of eight seed screens could help them vastly improve the speed and efficiency of their seed cleaning process. They can also be used to grind herbs.

Fancy Watering Can

A copper watering can, such as the Negishi one above, is probably the most beautiful option. I have found that most copper watering cans on the market are fairly small—0.5–2L—so they’re a better option for people with fewer/smaller plants, or for folks that like getting their steps in by walking to and from the sink.

I have a lot of plants to water, so I personally prefer watering cans in the 4–8L size. Haws, which has been trending a bit on social media, offers a nice watering can in that size range.

Books to gift to gardeners

I don’t know of any books on native plant propagation that I would heartily recommend, but Prairie Moon sells some books that might be worth checking out. Alternatively, if they’re a real freak, you might want to give them Carol and Jerry Baskin’s epically-long monograph on seed ecology: Seeds. Although it is my favorite book on seed ecology, most people probably wouldn’t want this as a gift. But if you catch your loved one on Google Scholar looking for academic articles on seed ecology, this might be a good fit!

Any book written by Eliot Coleman

Eliot Coleman is one of the best market gardeners, and I recommend all his books. The New Organic Grower is a classic, or if you’re curious about growing vegetables outside through the winter in northern climates—without heating—then you should read his Winter Harvest Handbook.

Most things by Charles Dowding:

Charles has an amazing—and beautiful—YouTube channel on no-till vegetable gardening, and he also sells a variety of books on the topic. I don’t agree with him on everything (he subscribes to some pseudoscientific beliefs about so-called “structured” water), but I can heartily endorse his gardening attitude and aesthetic.

Most things by Monty Don

Monty Don is the host of the popular British gardening show Gardeners’ World, and he writes mostly about ornamental gardens. Similar to Charles Dowding, he exudes a wonderfully laid-back attitude toward gardening that seems incongruous with the spectacularly well-tended state of his gardens.

Gift certificates:

As mentioned, gardeners can be hard to give gifts to because they often have very specific desires: a certain rare peony, a special rose, a hard-to-find native. So I’ll list below my favorite purveyors of plant material (seeds and live plants). If you know your loved one’s desires well, you could perhaps venture into ordering them a particular plant or seed, but otherwise a gift certificate is a safer bet.

Midwestern native plants and seeds

Windflower Natives

This nursery sells potted plants for many rare natives, and they are one of the only mail-order nurseries selling downy gentian, wood betony, or hoary puccoon. The owner is very communicative!

Prairie Moon Nursery

Prairie Moon has, as far as I know, the largest selection of seeds of any native plant nursery. I have complained about their incorrect/incomplete germination information, but obviously I have the utmost respect for folks devoted to making so many different natives available to the public.

Note that they are based in Minnesota. If your loved one lives in a different state, and cares about seeds of local genotypes, they may rather seeds from a local seller.

Vegetable and flower seeds (mostly non-native)

This is by no means an exhaustive list of seed providers, just a few that I tend to buy from:

Johnny’s

If you’re looking for reliable vegetable seeds, Johnny’s is maybe the best. Every seed I’ve ever planted from Johnny’s met or exceeded expectations. When I go to my local co-op, I usually recognize the various produce on display from Johnny’s catalog. If you’re looking for unique vegetables, this isn’t the place. But if you’re looking for reliable vegetables, this is it.

Truelove Seeds

Truelove is a great place to find some more unique varieties, and it’s a great company.

Uprising Seeds

Uprising is another great company. They have a huge selection of chicory/radicchio, which is probably my favorite vegetable crop.

Specimen plants (non-native)

Heirloom Roses

My preferred rose seller (I like own-root roses).

Solaris Farms

My preferred seller of rare peonies. Based in Wisconsin! Peonies sell for about $50 to $400 each depending on the variety.

Before we talk about the so-called “Empire Prairie,” let’s take a step back and try to imagine southern Wisconsin as it likely looked in 1830. In his 1914 History of Columbia County, J. E. Jones writes that “the most important land feature of the county is the high limestone prairie belt,” which he described as crossing through Green Lake, Columbia, and Dane Counties.1 Based on bedrock, elevation, and pre-settlement vegetation models, we can work backward from Jones’s description to identify the “high limestone prairie” he described.

First, let’s consider the adjective “high.” If we make an elevation map of South-Central Wisconsin, we can see that there is indeed a belt of relatively high land that extends southwest from Lake Winnebago. As this belt nears the edge of the Driftless Region (the area of the upper Midwest that never experienced glaciation), it meets up with a large blob of unglaciated highlands, which have a much rougher terrain characterized by steeper hills and deeper ravines. It therefore seems probable that the southwestern terminus of the belt Jones describes is the edge of the last glaciation, where the terrain changes dramatically.

Now let’s consider “limestone.” If we pull up a geological map, we can see that the bedrock of the previously identified highlands is largely of the Prairie du Chien group, which is dolomite with some sandstone and shale. There isn’t really much limestone in South-Central Wisconsin, so I’m guessing that when Jones says “limestone” he really means dolomite (which, like limestone is a sedimentary carbonate rock).

Finally, let’s consider “prairie.” LANDFIRE, a joint project of the U.S. government and The Nature Conservancy, produces a Biophysical Setting (BPS) map that models vegetation systems prior to White settlement based on physical site characteristics (e.g., soil, slope, aspect, climate) and past disturbance regimes (e.g., fire, flooding). Looking at the BPS map, there is clearly a belt of prairie along the dolomite uplands, connected by a matrix of savanna.

So that’s Jones’s “high limestone prairie belt.” But what about all the savanna that dominates the landscape?

A digression on savanna

I received a question from a reader asking about the difference between a prairie and a savanna, so let me briefly attempt to sort that out before I get any deeper into talking about “savanna matrixes.”

First, let’s recall what they have in common. Both prairie and savanna are fire-adapted ecosystems, with regular fires preventing the land from converting to forest. They are also both dominated by grasses like little bluestem, big bluestem, side-oats grama, and prairie dropseed.

As for what differentiates them, in simple terms, a prairie is a grassland with no trees, while a savanna is a grassland with sparse trees. John Curtis, a leading voice on the topic of Wisconsin vegetation communities, at times used an arbitrary cutoff of one-tree-per-acre: so if a community had one or zero trees per acre, it was considered a prairie, and if it had two or more, it was considered a savanna. But Curtis himself was critical of this cutoff, pointing out that when surveyors traversed Wisconsin in the 1830s, they recorded as savanna lands with just one tree per acre.2 Although we may be reluctant to trust 1830s surveyors with being the arbiters of the prairie/savanna boundary, consider that the Wisconsin savanna ecosystem is essentially exterminated, so those surveyors had the benefit of being able to actually witness savannas, whereas we can only speculate. It’s possible that these surveyors, like us, were just a little confused about where the prairie ended and the savanna started, but it’s more likely that they were recording actual changes in plant communities that we have limited ability to reconstruct today.

This is all to say that, while there were certainly areas of sparse trees across the area that Jones describes as a high prairie belt, the grass was essentially endless. And in light of the fact that all the treeless prairie existed within a matrix of oak savanna, it may not be ecologically accurate to consider prairies as discrete entities. Plants within the putative boundaries of a discrete prairie are not isolated from plants in a different section of prairie, and animals could certainly cross the savanna to get from one prairie to the other. But when reading historical accounts, it becomes clear that the difference between prairie and savanna had significant human impacts.

Hugh Jamieson, a Scotland-born settler of Wisconsin, chronicled his experiences settling on the high prairie belt. In addition to recording for us the shockingly low price of pork in the 1840s ($2.50 per hundred pounds), he also writes extensively on the prairie he encountered. In his account, the upland prairie—in contrast to the savanna—is an inhospitable place for humans. Fire, for instance, proved a frequent menace “sometimes if the wind began to blow a little, these fires would bound over the ground at a furious rate, and would sweep everything that stood in their way. Houses, stacks of hay and grain, and even live stock were often consumed by them.”3

In addition to being fire-prone, the prairie mostly lacked water. Narrating a stroll across the prairie, Jamieson writes “it was a tiresome, dreary journey, when performed on foot, to travel over this prairie in those days. Not even a drop of water was to be found except at a small pond, called the goose pond, near the center or about half the distance across.”4 I’ve taken the liberty of mapping Jamieson’s stroll across the prairie, and we can see that there was indeed very little water to be had within the prairie.

How the Empire Prairie got its name

The grassland that Jamieson perambulated would later come to be known as the “Empire Prairie,” but in Jamieson’s time it had many different names. These placenames are preserved in the names of churches, cemeteries, and towns: Arlington Prairie, Bonnet Prairie, Fountain Prairie, Spring Prairie, and Sun Prairie (see map below).

In what would eventually become the community of Leeds, settlers from New York named their area “Empire Prairie” in honor of their home state, which at that time was coming to be known as “the Empire State.” This title was memorialized in the name of Leeds’s general store—Empire Prairie Store—which stood at the main crossroads.

So with so many names, why did this vast 200-square-mile prairie come to be known only as the “Empire Prairie?” Martha S. Engel and Andrew W. Hopkins argued that the name became attached to the entire prairie when, in their 1916 Soil Survey of Columbia County, Whitson and his collaborators referred to it simply as the “Empire Prairie,” possibly influenced by the Empire Prairie Store and other businessed named in a similar fashion.5 I, however, don’t think this is the case (or at least, it’s not the full story). In the Western Historical Company’s 1880 History of Columbia County, the (unnamed) authors make two important observations:

1. They observe that the borders of the prairie have shifted in the 50 years since the original survey, writing “the prairies…are not exactly identical in extent now as when the country was first settled, for the reason that in some places, once prairie, there has been an invasion of timber growth,--this, in former times, having been checked by the annual prairie fires.”6

2. They describe the Empire Prairie and Fountain Prairie as two separate prairies, writing, “In the south central part of the county, the northern limb of Empire Prairie projects far out, like a great bay, through the southern and into the second tier of townships. In the southeast part of the county lies Fountain Prairie.”7 It would seem that, either due to woody encroachment or conversion to agricultural land, the northeast section of the prairie became separated.

It’s somewhat of a wild guess, but I’ve done my best to map out the boundaries of the prairie—within Columbia County—as described by the Western Historical Compant. If this is accurate, it makes more sense that Empire Prairie would be the chosen name, since the areas of Bonnet, Spring, and Fountain Prairie were no longer contiguous with the larger prairie (and possibly were no longer grassland). So it’s possible that “Empire Prairie,” as used in 1880, referred not to the entire 200-square-mile prairie as it existed in 1830, but to a much smaller prairie that included the town of Leeds.

However it happened, we inherited the name “Empire Prairie,” and it seems that’s the name we’re stuck with. Stay tuned for the last part of the series, where I discuss the remnants of this once-enormous prairie!

If there’s a seed ecology/prairie history topic you want me to write about, let me know by filling out this one-question survey:

Tell me what you want to read about