Testing hypotheses related to species size and light penetration
According to the size-advantage hypothesis, larger species are competitively superior to smaller species, mainly due to their ability to intercept light, shading plants below (e.g., Schwinning and Weiner 1998; Anten 2005; Tracey and Aarssen 2014; Tracey et al. 2017). Consistent with this, we found that light penetration significantly decreased with mean plot height (Fig. 5). Importantly, the height of species in sample plots, as measured here, left approximately 90% of variation in plot-level light penetration unexplained. Some of this unexplained variation is inevitably due to the way in which we estimated plot height – future assessments based on measuring the heights of all plant species in plots would be of benefit but were not possible for this study. Regardless, this result highlights that while tall species have the capacity to shade smaller plants, the biological significance of this advantage may be greatly reduced by the size distributions of tall species in plots, and by other factors that contribute to variation in composition within herbaceous plant communities. Inevitably, other factors, like the available soil nutrients, can limit the height of the canopy by controlling which species can persist (Borer et al. 2014). Large species with large minimum reproductive size thresholds (Tracey and Aarssen 2019) may be less likely to persist in low soil nutrient settings and self-thinning among tall plant species may create regular gaps, allowing light to reach smaller species (Schamp and Aarssen 2014). Thus, nutrient limitation and competition among tall species may decrease the abundance of large plant individuals per area, simultaneously increasing light penetration and altering the importance of light as a previously limited resource. Herbivory can also impact canopy development directly, and indirectly by influencing plant investments in herbivore defense.
We found evidence that small species abundance and richness are limited by light penetration; these patterns were consistent when we considered reproductive plants and small species were defined as those below the median height of species in the community, but inconsistent when small species were defined as those in the first quartile of species heights in the community. These findings were less consistent across both definitions of small species when we did not focus on reproductive plants (Table S1). Importantly, the coefficients of determination (R2 ) for regressions between light penetration and our response variables, when significant, ranged between 8-11% (Fig. 6). Furthermore, only 4.2% of the variation in the composition of small species, defined using the median, was explained by light penetration (Fig. 7). Low light penetration had a significant, but meagre impact on small species abundance and richness in this community.
These results, in combination with the above results showing that large species contribute only slightly to patterns of light penetration, provide useful context for understanding the conundrum of small species diversity in natural systems in spite of experimental observations that small species are at a competitive disadvantage. First, light penetration in herbaceous systems is relatively low, but less so earlier in the growing season (i.e., June in this study system). It is possible that some small species take advantage of higher light penetration early in the growing season. This is consistent with observations that smaller plant species often flower earlier in the season (Du and Qi 2010; Sun and Frelich 2011; Segrestin et al. 2020). Second, within-plot variation in light penetration, reflecting heterogeneity in plot canopy structure, had no impact on small species abundance and richness, suggesting that local sunflecks are relatively unimportant in maintaining small species in herbaceous communities. This may be because plots with more variation in canopy structure tended to be less shaded on average, reducing the role of sunflecks as a critical source of light. Third, the fact that small species richness is only modestly impacted by light penetration suggests that many small species possess some mixture of shade avoidance and shade tolerant strategies that allow them to persist in shaded conditions. Shade tolerance may contribute to increased reproductive economy in smaller species (Aarssen 2008). Shade tolerant species implement morphological and physiological traits which increase efficiency of carbon gain and stress tolerance, allowing them to persist under low light conditions (Valladares and Niinemets 2008). Additionally, small species richness, defined using the 1st quartile, was no lower when light penetration was lowest, indicating that the richness of the 25% smallest species in the community was maintained under increasingly low light availability. Adaptation to low light levels via phenotypic plasticity in morphological traits may explain the persistence of small species richness under low light (Ballaré 1994; Urbas and Zobel 2000; Hallik et al. 2009; Niinemets et al. 2015), as some individuals may express traits associated with greater light capture (e.g., high specific leaf area; Valladares and Niinemets 2008), thus maintaining species presence under reduced abundance. The characteristics of the light environment in our community, and likely the characteristics of many resident small species ensure that any disadvantage they suffer in competition for light is minimal.
The effects of shading on small species abundance and richness were more consistently observed and stronger in general when we focused on reproductive individuals in sample plots. This demonstrates that although ‘median’ small species were able to persist under low light with no detected loss in abundance or richness, some were not successful enough to reach reproductive maturity. The presence of flowering individuals in a sample plot is evidence of a species’ capacity to successfully establish in that environment. Our results make it clear that light penetration does indeed affect which small species can find success, but its impact on small species abundance and richness is relatively minor (8-15% of variation explained). Our results also support the contention that focusing censuses on flowering species when studying coexistence, can clarify plant community dynamics (Schamp et al. 2016, Schamp and Jensen 2019).