Iowa is the most altered state in the U.S. Nowhere have natural communities been more thoroughly destroyed. I started thinking about ecotypes when I was in my early twenties collecting seeds on little “postage stamp” prairies in Iowa. Ecotypes can be thought of as plants from a particular geographic area that tend to share similar genes and also observable structural characteristics (morphology), though often the concept is simplified to geographic sources, regardless of the magnitude of any real differences between plants. For that reason, I’ll use “ecotype” and “source” interchangeably here.

Often restoration and reconstruction projects for natural communities limit themselves to local ecotypes, which often means, on the conservative end, using only seed gathered within the immediate vicinity of a project area, or on the more liberal end, sourcing seeds from within a county, adjacent counties, or perhaps up to 100 miles or so away.

Myself in youth using a leaf blower in reverse to suck up and pulverize ground plum (Astragalus crassicarpus) pods on an Iowa prairie. For those with PPE concerns, I will add that this was a staged photo.

The seeds I was harvesting on those Iowa prairies were either going directly to market after cleaning, or they were going into restorations or plantings, which would subsequently be harvested for use in other restoration projects. As I moved from one small and geographically isolated population to the next, the following questions increasingly weighed on my mind:

1. What is missing? Often the bags strapped to my waist were heavy with seeds, but what about genes? Would the ghosts of prairies past in the surrounding seas of corn and soy haunt our success?

2. How good of a match are native populations in remnant prairies, more often than not located on marginal, rocky or sandy ground, for prairie reconstruction projects located in cool-season pastures or cultivated fields? Such sites might be geographically close, but they are often situated on different soils, which have seen their structure deteriorated and their nutrients and microbial communities altered by more than a century of agricultural use.

Typical postage stamp prairie surrounded by horizons of corn in western Iowa.

3. What about climate change? Plants on those postage stamp prairies surrounded by miles of corn and beans are going nowhere fast. Already, temperature and precipitation patterns have changed enough in much of the Midwest and Great Plains that you would have to travel a hundred miles or more to find the same/similar climate in terms of average temperature and precipitation at the time of European settlement, and that ignores the extremes of our new climate. Still, we often use climate to make guesses about how to define ecotypes. Not only have native plant populations passed through extreme genetic bottlenecks, but the prairie landscape has been fragmented to such an extreme degree that if there were other suitable sites on the landscape, the plants would not be able to disperse their seeds to them. Not even Donald Trump could build a better wall than what an 80- or 160-acre Iowa cornfield is to a compass plant. Note: Donald Trump never crossed my mind once back in those days. It was an age of reason.

4. Given these things, how much does the ecotype concept, particularly when interpreted narrowly, really help us?

Here is a brief synopsis, admittedly simplistic, of what we know. Plants from different locations, whether separated by altitude, latitude, longitude, etc. differ in terms of both their structures and genes. The timing of flowering and growth (phenology), the size of all or parts of individual plants, and tolerance of extremes of heat or cold are all things that can vary. We know this, in large part, due to common garden experiments. What we observe when we look at a plant is the result of genes and the environment. If plants from locations A, B, and C, are all grown together in one place, it can be assumed that their differences are genetic and not based on environment alone , because the environment is held constant. This ignores epigenetics, which could be very important, but methods to suss out its effects haven’t really been developed as far as I know. Phenological events like flowering time and time of emergence can affect how plants interact with consumers (e.g. pollinators) and one-another. A classic scenario is the one where a flower emerges earlier due to a warming climate, but its pollinator is not able advance its emergence to the same degree, leading to an asynchrony that harms both species. In such a case, environmental conditions are the cause. However, use of an ecotype that differs in phenology from a local source could cause similar asynchronies.

A monarch butterfly visits meadow blazingstar (Liatris ligulistylis).

For those reasons, “local is best” is the mantra, if not the dogma, of many in the conservation and restoration community. It’s a cautious and logical approach. However, in recognition of the altered and changing world we live in and recognizing that ecological restoration and reconstruction projects are  essentially the only way many species can move across today’s fragmented landscape, I would urge people to be a little more liberal in how they apply the ecotype concept.

I will start by briefly summarizing my research on the topic. I was not only concerned with whether plants from points A and B were a little different from one-another. I was concerned with whether or not any differences they might have actually affected how well they established in a realistic prairie reconstruction setting. Additionally, I was not concerned with one species, but with a dozen, all of which I sourced from three distant locations (one in Nebraska, one in Kansas, and one in Oklahoma). Thus, I was planting twelve species together, all sourced from the same location  (a twelve-species, single-source seed mix). The mix from Nebraska was sown next to a mix with the same species in the same amounts from Kansas, and another from Oklahoma. In turn, at each site from which I sourced the twelve species (NE, KS, and OK), I established common gardens of prairie reconstructions so that differences in outcomes would be due largely to the seed sources rather than the particular environmental conditions they were grown under.

What did I find? There was no consistent local advantage. Sure, there were cases where the local source performed best (measured as species-specific biomass 1,2, and 3 years after planting), but those cases were no more common than one would expect from chance. Sources often performed similarly, and in some cases non-local sources performed best. In addition, seed source did not affect the number of sown species that ultimately established 1, 2, or 3 years after planting. Two articles from this research are available open access at:

Seed Source and Prairie Establishment

Seed Source and Prairie Seedling Survival

Stiff goldenrod (Oligoneuron rigidum) seedling marked and tracked for survival on a research plot in Nebraska. There, the local Nebraska source did have the highest initial seedling density, but there were no differences in the survival of emerged seedlings between the Nebraska, Kansas, and Oklahoma sources.

Now, is that work conclusive? Absolutely not. It followed reconstructed prairies for three years, and it used only a small subset of possible species, sources, and growing sites. However, its results should give pause. It illustrates how focus on one species or one site could have led to false support of a general local advantage. In addition, initial establishment is critical for many prairie species in a restoration / reconstruction setting.  In perennial plants, the most important demographic parameter for maintaining stable populations is usually  year-to-year survival rather than seed production, so getting plants growing is the most critical step. Initial establishment also impacts the extent to which invasive species can get a toehold.

Was I surprised the results? No. The plots that were seeded were different from even the nearby prairies from which the seeds were sourced, our climate has changed, and the genes I gathered as I collected seeds from remnant prairies, even though they were  big prairies with fairly large populations of many species, were most likely just a subset of what was once available in each region.

Stiff goldenrod in its second year from seed on one of my research plots in Kansas. In Kansas, the Nebraska source for stiff goldenrod established better (grams biomass / square meter) than the local Kansas source and the Oklahoma source.

What does other research say? There is very little work that goes beyond whether or not plants have genetic or other measurable differences by taking the additional step of quantifying how well they actually perform in a realistic setting. However, similar work, which focuses on only single species or a few species, is mixed or equivocal on local vs. non-local sources.

Local Sources VS. Cultivars (Paywall)

Local VS Non-Local Along Longitudinal Gradient (Paywall)

That said, we still know that ecotypes can differ in measurable traits, including phenology, and that has the potential to be a big deal.

Let me return to my suggestion to conservation and restoration professionals to be more liberal about ecotypes. The ecotype concept was developed under a view that assumes stasis in the natural world, but everything these days is rapidly changing. Certainly, when local sources are available, they should be used. However, in order to conserve biodiversity, we should also look more regionally, take sources that may be geographically more distant, and use those in combination with our local sources. I’m not saying that that a project in Wisconsin should use seeds from Texas, but it may be worthwhile to consider sources from Iowa or Illinois. Below are the reasons I find most compelling.

1. The effects of source location vary by species such that there is no one-size-fits-all rule or relationship for sourcing plant material.

2. If any one source performs poorly, it will tend not to survive and reproduce, and it may then be compensated for by better performers in a mix that includes a variety of sources.

3. Ecological asynchronies can be caused by sourcing geographically distant plants, but they can also be caused by climate change. It is not, therefore, clear that including distant sources that differ from local sources in phenology  will, with the backdrop of climate change, always lead to asynchrony. In fact, when different sources are mixed, a broader phenological range (e.g. more variation in flowering dates) might reduce asynchrony.

4. Best practice in restoration/reconstruction is to utilize as many native species as appropriate and practicable, which often diminishes the importance of any one species through redundancy. This increases community resistance to and resilience from disturbances. Since many specialist consumers specialize on genera or families rather than species, potential asynchronies in phenology can be minimized by including multiple species from families and genera that are native to the same region but that grow or flower at different times from one another, thus extending the season.

5. A regional perspective provides the seed buyer more options in the market, both to locate purveyors of desired species and to control costs, both of which can be leveraged towards creating a seed mix more diverse in species, which has known benefits.

Drought parches Konza Prairie in 2012 after a spring of unprecedented warmth that caused big bluestem (Andropogon gerardii) and heath aster (Symphyotrichum ericoides) to flower in May. Fluke, or more to come?

6. There is simply no other way for many species to move over the landscape as quickly as they need to in response to climate change. It’s a forgone conclusion that natural communities are going to profoundly change; some already have. The contemporary fight is to maintain biodiversity. Unfortunately, we don’t know enough to confront this task with clarity, but we also don’t have time.

At the very least, people should think more critically their sourcing procedures in the narrow context of single projects and the broad context of regional biological and genetic diversity. This is not a topic where reality lends simple, repeating, and unthinking procedure.