Back during the winter of 2009-2010 I sowed restoration plots that consisted of twelve grassland species at three sites in three states. At each site I sowed the same species and the same weights of live seeds for each species, but I manipulated seed source at the community level. The general design of the experiment is shown in Fig. 1 below.

Figure 1. Design of common garden showing the list of species sown on to each plot (left) and representative blocks (right, n=4 at each site). Mean temperatures and precipitation reflect 1981-2010 climatology at nearest weather stations.

The purpose of the experiment is to determine whether seed source effects community and ecosystem level properties. We know that provenance affects the performance and expression of traits for individual species at single sites, and these responses are idiosyncratic, meaning that sometimes local sources perform best and sometimes other sources perform best at the species level. What we know less about is whether or not we see source differences for species when they are grown in a more realistic community setting, whether source effects are consistent across sites, and whether effects scale to community and ecosystem level properties. Also, there is a mixed treatment, which should give insights into whether variation within single sources is limiting.

In this experiment, terrible drought in Oklahoma led to very poor establishment in the first year (2010) and near complete die-off in 2011, so I exclude that site below and show some responses of different sources at the Nebraska and Kansas sites. I focus on data from September harvests of aboveground biomass for all sown species combined and non-seeded species combined.

I will say without showing any evidence that the individual species differ from one-another in their patterns. Together, these have led to source effects on both seeded and non-seeded productivity (Figure 2). But I thought the pattern in Figure 3 was especially cool, and I’m looking forward to adding to it next month

Figure 2: This shows seeded and non-seeded biomass grouped by seed source (x-axis) with values from the NE site in dark grey and values from the KS site in light grey. Establishment of seeded biomass was similar and relatively high among sources in NE compared to Kansas. In Kansas, the NE source had the highest seeded biomass (excluding the mixed source). Generally, non-seeded biomass is negatively associated with seeded biomass as the next figure shows as well.

Figure 3. Here biomass for all KS plots is in red and biomass for all NE plots is in black. Small points are values from 2010 and large points are values from 2011. I think this looks kind of threshold-ish, so that’s pretty cool.