Intraspecific variation mediates density dependence in a genetically diverse plant species

Abstract

Interactions between neighboring plants are critical for biodiversity maintenance in plant populations and communities. Intraspecific trait variation and genome duplication are common in plant species and can drive eco-evolutionary dynamics through genotypemediated plant-plant interactions. However, few studies have examined how species-wide intraspecific variation may alter interactions between neighboring plants. We investigate how subspecies and ploidy variation in a genetically diverse species, big sagebrush (Artemisia tridentata), can alter the demographic outcomes of plant interactions. Using a replicated, long-term common garden experiment that represents range-wide diversity of A. tridentata, we ask how intraspecific variation, environment, and stand age mediate neighbor effects on plant growth and survival. Spatially explicit models revealed that ploidy variation and subspecies identity can mediate plant-plant interactions but that the effect size varied in time and across experimental sites. We found that demographic impacts of neighbor effects were strongest during early stages of stand development and in sites with greater growth rates. Within subspecies, tetraploid populations showed greater tolerance to neighbor crowding compared to their diploid variants. Our findings provide evidence that intraspecific variation related to genome size and subspecies identity impacts spatial demography in a genetically diverse plant species. Accounting for intraspecific variation in studies of conspecific density dependence will improve our understanding of how local populations will respond to novel genotypes and biotic interaction regimes. As introduction of novel genotypes into local populations becomes more common, quantifying demographic processes in genetically diverse populations will help predict longterm consequences of plant-plant interactions.

Spatial models can improve the experimental design of field‐based transplant gardens by preventing bias due to neighborhood crowding

“…Growth model parameters were based on the subspecies‐cytotype group differences in growth rate and plant tolerance to neighborhood crowding proposed by Zaiats et al ( 2021 ). Neighborhood crowding represents a cumulative measure of per capita effect on a quantitative trait from the immediate set of neighbors, corresponding to a measure of competition response by target plants (Chu & Adler, 2015 ; Goldberg & Fleetwood, 1987 ).…”

“…Neighborhood crowding represents a cumulative measure of per capita effect on a quantitative trait from the immediate set of neighbors, corresponding to a measure of competition response by target plants (Chu & Adler, 2015 ; Goldberg & Fleetwood, 1987 ). The statistical growth model quantified intrinsic growth rate and crowding effect following Equations 1 and 3 (detailed below), and we refer readers to Zaiats et al ( 2021 ) for further details on parameter estimation. Briefly, the common garden design allows for 11 distance‐to‐focal plant treatments within a 4 m neighborhood radius, while the non‐linear crowding kernel can accommodate a range of spatial neighbor effects, from a sharp to gradual declines in neighbor effect with distance.…”

“…Briefly, the common garden design allows for 11 distance‐to‐focal plant treatments within a 4 m neighborhood radius, while the non‐linear crowding kernel can accommodate a range of spatial neighbor effects, from a sharp to gradual declines in neighbor effect with distance. Models for plant demography in this common garden indicated different responses to neighborhood density based on subspecies identity and ploidy (i.e., cytotype) variation (Zaiats et al, 2021 ). Specifically, tetraploid variants demonstrated more conservative growth and greater tolerance to neighbors than their diploid variants.…”

“… where is the predicted growth of plant i belonging to group s with k neighbors, with the superscripts ( F ) and ( B ) corresponding to the full and base models, respectively, and is the unexplained variation. The parameters c and b in Equation 3 determine the shape of the neighborhood effect, from sharply to gradually declining with distances, which was empirically estimated to near negligible effect beyond the distance of 2.5 m away from the plant (Zaiats et al, 2021 ). Similar models are widely used to quantify the effect of spatial interactions on plant fitness (Adler et al, 2010 ; Hülsmann et al, 2020 ).…”

“…We calculated the scenario under which the 95% credibility interval (CI) correlation Pearson's coefficient was r > .99, indicating that impacts from spatial interactions among plants are unlikely to impact inference on mean group differences. The empirical estimates of plant growth using spatial and non‐spatial models are detailed in Zaiats et al ( 2021 ) and Richardson et al ( 2021 ).…”