email alerts

To receive email alerts for new posts of this blog, enter your address below.

Friday, October 12, 2018

Giving Rare Seeds a Good Start

More Than Perhaps You Wanted To Know
(but perhaps less than you need to know)
If You Want To Competently Restore An Ecosystem

Rarely Asked Questions (RAQ)
and Possible Answers (PA), not pretending to be the last word – 
in this evolving discipline of the restoration of quality plant communities

RAQ1: Why is it important to plant seeds in the right place? 

PA: Some people say we wrongly “play God” in deciding which seeds to plant where. And in theory, the most natural method would be to plant all species everywhere – and let God or nature decide which would prevail. But our most important seeds are rare – and harvesting rare genotypes from the wild, as we do, diminishes the places they’re taken from. Take too much too often – and we do damage. This is especially true of the seeds of now-rare, formerly dominant conservative species. 

So, we don’t want to waste them. Also, the seeds of wetland plants won’t propagate on dry ground, and although middle-moisture seeds might survive for a while in somewhat drier and somewhat wetter habitats, they’d be replaced by more fit species in time. The same could be said for species of full sun or deep shade. Thus, if our goal is to conserve rare biodiversity, we can’t ethically waste. And we don't want to waste our time.

Reassuring note: This account – drawn mostly from experience at Somme Prairie Grove – won’t concern itself with the complexities of less common soils like sand, clay-pans, etc. but will stick with prairies, savannas, and woodlands rising from the typical black soil of the tallgrass region. 

RAQ 2: Should we plant the same species in bare ground (where brush was cut) … compared to a grazed-out old field that just needs more diversity?

PA: Use different mixes. This is an important question, worth being aware of, but the answer is complicated. Many of the most important species won’t do well on bare ground, so we save them for “turf” mixes. But detail on this tangled question is relegated to Endnote 1.

RAQ 3: How do I recognize where I should stop broadcasting the prairie seed mix and start doling out savanna seeds?

PA: The first answer is that you’ll want to overlap those mixes a bit – as we’re just not that smart as to know for sure, and, indeed, the species mix naturally. 

The second answer is that it depends on how far apart and how big the trees are in your savanna. A rule of thumb: if full sun shines on a piece of ground for two thirds of the day or more, plant prairie seeds. If full shines from one third to two thirds of the day, plant an intermediate mix. If full sun is less than one third of the day, in an oak woods, sow an oak woods mix. But there’s much more on this, below. 

RAQ 4, and last: In restoring an oak woodland, should all the invading shade be removed before I plant my first seeds, or should the process be staged? 

PA: It has worked best for us to keep much oak shade in early stages. But our knowledge of how to restore biodiverse woodlands is even more primitive than with prairies and savannas. Some approaches lead to an understory of briars, poison ivy, and such rank vegetation that seems to head in the opposite direction from quality. We're only in early stages of the experiment of opening the canopy enough for oak reproduction, planting diverse species for that level of sunlight, and nursing that intensive-care-needing community (by mowing, scything, weeding?) enough to keep the more aggressive species from diminishing or preventing the establishment of a diverse conservative, stable turf. (More on this in Endnote 5.) 

Some Visuals That Might Help

Many restoration principles can be expressed simply, but recognizing how to apply them on the ground in a specific place is the real challenge. Don’t expect to be perfect at it. In fact, just go ahead and assume that you won’t be right all the time and in every detail. But thinking about the principles ought to help. And field exercises help the most. The visuals below were designed to help prepare for field exercises. 


The graphic above seems to say: “Put the Prairie Mix in the prairie and the Savanna Mix in the savanna.” No, sorry, that’s not the way.  

Probably the seed mixes shouldn’t be called “Prairie,” “Savanna,” and “Woods.” 
Probably they should be called something like “Full Sun,” "Intermediate,” and "Dappled Shade.” 
For more on this unfortunate terminology, see Endnote 3.

This graphic gets us closer to reality. Most open savanna should be seeded with the “Prairie” mix. But immediately under and around and especially to the north of isolated large trees (considering how the sun’s rays slant at our latitude), plant the “Intermediate” mix. In the darkest parts of the savanna, blend in some “Woods” mix. 

There is no “maple forest mix” on this map, because the North Branch had little of the maple/basswood or maple/beech forest (and those fine natural communities are a good deal less threatened than oak communities, so there’s been less research as to restoration techniques or conservation goals or priorities). The restoration we’re considering in this exercise is for oak woodland and oak forest, which have much brighter understories than a maple forest. 

Notice, above, that the seed mixes to be planted in a savanna may be mostly the Open (“prairie”) mixes, and a woodland may have open meadows and many areas (depending on the spacing of the trees) for the “Intermediate” mix. In fact it is in these areas that the bur and white oaks reproduce. 


This final graphic is “as good as we’ll get” in this post – and perhaps a “good enough” approach for this simple site. It adds wetness to shade in our decision-making. On the North Branch we deal with few areas of “dry” soils or even “dry-mesic” soils, so we’ll focus here on “mesic” and “wet-mesic” soils. 

(Note on the obscure word “mesic.” It’s just fancy jargon for “medium” moisture. That is, it’s half way between dry and wet. I apologize for it. But it’s so widely used in conservation, that we may be stuck with it. See Endnote 3, in the unlikely event that you want to think more about jargon considerations.)

The graphic above was made in 2018. By 2021, we've decided to divide the very broad "Intermediate" category into two parts. Thus we now make and sow eight different seed mixes in an area like the one shown above:

MP: Mesic Prairie  
MBI: Mesic Bright Intermediate
MDI: Mesic Dark Intermediate
MW: Mesic Woods  
WMP: Wet-Mesic Prairie 
WMBI: Wet-Mesic Bright Intermediate
WMDI: Wet-Mesic Dark Intermediate
WMW: Wet-Mesic Woods  

How do we define our shadiness terms? Prairie needs full sun nearly all day long. Woodland has only dappled light. The "Intermediate" areas have full sun from roughly two-thirds to one-third of the day – whether because one big tree puts it in full shadow for part of the day, or because the overall tree canopy has enough holes in it that any given piece of ground gets sun for roughly 80 to 20 percent of the day. 

The graphic may seem complicated, but on the ground it’s not so bad. You’ll find it quickly becomes second nature to check where you are in relation to trees: a) far from, b) near, and c) under. Recognizing whether you’re in mesic or wet-mesic or wet soils is trickier. You can often tell from the existing vegetation. (See Endnote 4.) For example, white oaks indicate “mesic” and swamp white oaks indicate “wet-mesic” or “wet”. But you can also often tell (especially as you’ve become more familiar with your site) because on the day you’re seeding, the mesic areas will be light brown and the wet-mesic will be dark and damp. 

Also, a person who’s studying this insanely closely may notice that the “LESS WRONG” graphic treats some areas differently from the “BEST FOR NOW” graphic, as to whether an area deserves “Prairie,” “Intermediate,” or “Woods” seed.  

One Last Point 

When broadcasting seed, consider competition from the species already growing there – unless it’s bare ground where brush was just cut.  If an area is dense with tall goldenrod and briars, don’t invest too much rare seed there yet. With good management, the vegetation should get increasingly conservative, and as that happens other, more-conservative species can be seeded in. 

The highest-quality seed mixes should be broadcast into thin old-field turf (made up, for example, of bluegrass, daisy, carrot, and early or gray goldenrod). This approach has been proven by long experience on the North Branch, Poplar Creek, Spring Creek, Flint Creek, and Orland preserves. Check out photos of and comments on such areas.

What’s your experience?

If you do this work, please keep careful records and share what you learn. From time to time we incorporate various people's experiences in posts like this. 

Endnote 1

Inter-seeding deserves more than an endnote, but this most-important and best way to plant diverse conservative seed at least deserves a quick summary here. 

For high-quality mesic prairie seed, the most receptive is a diverse turf of “old field” species like Canada bluegrass, poverty oats, timothy, wild carrot, and ox-eye daisy. Even if you don’t know these species, you can probably recognize this kind of area as one where you can see through the plants down to soil all summer long. If the shade is too dark, prairie seedlings will die.  

In sandy soil at Nachusa, the stewards recommend planting all the species with the first seeding, and that seems to work there. At Fermilab, Dr. Betz recommended starting out with “first wave” species like big bluestem, Indian grass, and prairie dock – adding conservatives later. Most people today don’t recommend that approach.

At Somme we have mostly used two sets of mixes, one for bare ground where brush was recently cleared (although we often sow no seeds the first year to give us time to spray out invasive seedlings and re-sprouts). Sometimes in such situations we sow non-aggressive grasses that first year, to hold the soil and allow us still to spray broad-leaf herbicides). 

The second major mix is for higher-quality established turf. In an “old field” of daisies, bluegrass, poverty oats, and even smooth brome, we get good results from broadcasting the very “highest conservatives” including prairie clover, dropseed, Leiberg’s panic grass, shooting star, the phloxes, and everything good. Over the years, they out-complete and replace the weeds. (However, if the bluegrass or brome is so dense that it heavily shades the ground by summer, it is necessary to burn it for a year or a few before planting.)

In bare-ground woodlands, we hold back on such species as hepatica, betony, shooting star, and rue anemone until there’s a turf to hold them. We have less experience with woodlands.

In rank areas of tall goldenrod, sawtooth sunflower, Canada thistle, etc. we suspect there’s an appropriate mix to break promote diversity, but we’ve done little careful research on alternatives. In such areas, leadplant, dropseed, prairie dock, and Culver’s root (among many others) succeed well.  

Endnote 2

Fallen tree leaves act differently from the previous year’s old-field vegetation. We have found that few seedlings of most species can make it up through dense woodland leaf litter. So we try to plant such areas only after the leaves have been burned off. In some years, wind blows leaves into piles and leaves some areas bare, so we then seed those bare areas. 

Diverse woodland understories seem readily established in burned areas if sufficient seed can be broadcast. Mesic species that do especially well for us (and seem especially compatible with other conservatives) in the early years include elm-leaved goldenrod, golden Alexanders, nodding fescue, starry campion, and wide-leaved panic grass. Prime wet-mesic species include zigzag goldenrod, great blue lobelia, Virginia rye, and many sedges. Indeed, sedge species seem to foster diversity in woods from dry to wet. Since our seed teams can’t always identify all the sedges, we ask that they just gather what they find and label the bag with the habitat (wet prairie, dry open woods, etc.).  

Endnote 3

Conservation needs public support. Jargon alienates people. 

I tried to get the annoying word “mesic” out of this summary. Why not just speak English and say “medium”?! But in this case, that leads to the seed mix called “Medium Intermediate” – which is ridiculous. 

If anyone thinks it’s worth the challenge of trying to clarify terminology, I suggest possible alternate names:
For wetness: Wet, Wettish, Medium moist, Dryish, Dry.
For canopy cover: Full Sun, Half Sun, Dappled Shade, Deep Shade (or Full Shade).
Or perhaps: Full Sun, Part Sun, Light Shade, Full Shade.

(Note: Deep Shade or Full Shade would apply to maple forest. On the North Branch, we don’t have such a mix.)

It’s good for all stewards to know what “mesic” and “wet-mesic” means, so  you can read the technical literature. But we also need language to use with the general public and new volunteers.

Endnote 4

Since the quality species you’re seeding are likely not already present, you often you have to judge the wetness of an area by the weeds or invaders. 

Indicator plants for seeding mesic prairie:
Ox-eye daisy
Wild carrot
Canada bluegrass
Early goldenrod

Indicator plants for seeding wet-mesic prairie:
New England aster
Sawtooth sunflower (but don’t plant if it’s too dense)

Indicator plants for seeding wet prairie:
Swamp milkweed
Blue vervain

In Intermediate areas, there’s rarely any vegetation beside brush. Check the Intermediate seed mix lists to find possible indicator species, if any are present. 

In woodlands, often the spring ephemeral species survive.
Indicator plants for seeding mesic woods:
Prairie trillium
Cut-leaved toothwort

Indicator plants for seeding wet-mesic woods:
Swamp buttercup
Fringed loosestrife
Mad-dog skullcap

Indicator plants for seeding wet woods:
Cardinal flower
Swamp buttercup

Endnote 5
Our oak woodland restoration has worked most dependably when we've left enough oaks to cover the ground with sufficient leaves each fall to make for a good burn. For better or worse, that means too much shade to allow oak reproduction. Maybe that's fine for now, as we don't need oak reproduction in a woods with that many oaks. They can start reproducing again when the old trees die and allow more light. 

But the problem is that such a light level is insufficient for many of the conservative species that we hope to conserve sufficiently that they'll end up in a community that has enough light for oak reproduction. In fact, sustainable oak reproduction may depend on such a turf of conservative species. In some cases we've watched old oaks die only to be replaced by dense stands of non-diverse, rank vegetation (woodland sunflower, Joe Pye weed, briars, etc) making so much shade that no oak reproduction occurs. The oaks reproduce in grassy turfs on the edges of the groves but not in recent openings. Perhaps what we're looking for takes more time (many decades?) or perhaps it takes more diverse species than we have? or perhaps there are other answers. We're experimenting. We'll report any results. We're always interested in the results of others. 


For detailed photos about sowing seeds in the "prairie" end of this spectrum, the one we know best, check out


  1. It is VERY important to plant seeds of conservative species in suitable places, because, as noted, the seeds are hard to come by and taken from a place at which the species has been successful. Restoration should be amplifying desirable native seeds – each seed collected should produce MORE than one seed at the restoration site so excess can be used a new restoration. If collected seed is not amplified, collecting natives is contributing to degradation of the broad environment, ie is unethical. Attention to quantities collected, dispersed and produced in the restore area is important to progress in increasing efficiency/amplification. Gardens producing seed of rare species are also helpful in amplifying seed. Plants in gardens have capacity to produce hundreds of times as many seeds as wild individuals.
    A BETTER strategy for rare natives than MIXES is to treat each species separately. Species segregate according to temporal patterns of light and moisture integrated over many years and history. Populations occur in patches with broader communities. Neighbors/associates are a way people can access moisture patterns which are not based on moisture measurement by instruments. This idea is intrinsic in the idea of ‘indicator’ species but more specific. (The associates lists in S&W 1979 & 1994 are much better for this purpose than those in W&R 2017 because the latter’s lists are longer and include weaker associates.) One will achieve more success per seed if one uses experience and knowledge (preferably from a dozen sites) to match each species being dispersed to a particular place (finer than a mapped community). Additionally, the individual species approach builds patches of populations rather than homogeneous distribution of mixes. Admittedly, the individual approach does take a lot more person·hours than mixes, but I believe the time is well spent.
    Many species are ‘comfortable’ in multiple ‘communities’, others are more specific to particular local conditions that may be rare within a named community. Categories (such as communities) are necessary for writing, but if they don’t lead to knowledge ‘on the ground’, they are best left in books. The seeds are precious and only people that are willing to learn should handle them. The pictures of children collecting, or dispersing seed may make reader/viewers feel good, but inexperienced youth are unlikely to be useful for conserving natural communities.
    My grandson went to “Central middle” school which struck me as redundant. I have no problem with ‘mesic’, but ‘average’ will probably be more understandable than ‘medium’, if you try to make a change. For weather I use: Sunny, Mostly, Partly and Cloudy & Rain. Five categories are memorable and useful, but nature is much more complex. Encourage volunteers to understand nature deeper than communities. New volunteers need to learn jargon if they are to continue to participate. I suspect it is harder in the age of ‘twitter’ but getting a quality volunteer who wants to learn probably is more valuable to nature than ten times as many ‘medium moist’ volunteers.
    I find it disheartening that your ‘mesic indicators’ include non-native species. Ox-eyed daisy and wild carrot have continued to be abundant at Somme Nature Preserve. I was regularly in Glenview from 1999 to 2013 and visited Somme NP each year during that period. From what I remember, ox-eyed daisy abundance was increasing, and the species was not being displaced by prairie conservatives. Cleared and herbicided areas do take a long time to recover to a native community, but I wonder if using non-natives as ‘indicators’ has encouraged their persistence. I hope you are successful in displacing non-native species.

    1. Interesting comparison between Swink and Wilhelm and Wilhelm and Rericha volumes - thank you! The insect associates listed in the latter volume make it well worth the cost. I do miss the basic morphology illustrations of the previous volume but perhaps the authors felt that information was easy enough to find online these days so it wasn't repeated in the second book?

    2. Thanks, JWPboss, for wise observations. I have a response to your good thought on whether to restore rare species in mixes or individually. I agree with all you wrote, but there is an additional consideration. Yes, we carefully research and restore rare species in what seem the best 'micro-niches' if we have only a few seeds. But, as you recommend, we often take pains subsequently to produce large numbers of seeds of these species, by protecting them with exclusion cages, or pampering them in wild gardens. In the case of forked aster (Aster furcatus), we first successfully established a few small populations, carefully harvested the seed, and put some of it into our wet-mesic woods, wet wood, wet-mesic intermediate, and wet-intermediate mixes. Our reason for that is that we just didn't think we were so smart that we could figure out all the best spots. And it was like a "free bonus" to have the seed dispersed widely with the other seed. We were very impressed to see how many and varied places supported new populations of forked aster. We expect that many of those populations won't persist, but we expect the same of those populations that we carefully researched and sites. In fact, we wouldn't be surprised in all of our (how hundreds) or little populations fade away - or if all but a few fade way, and those few turn out to be in places we would not have expected. We would also not be surprised if the genetics of the species went through ongoing transformations as it dealt with the new habitat, new climate, and other changes ... and perhaps will survive at Somme specifically because it had such widespread opportunities for adaptation. (For that purpose we also try to include seeds from as many sites and habitats as possible.) This is not the only good approach, but we believe it is one good approach.