Density‐dependent recruitment structures a heterogeneous distribution of herbivores among host plants

Abstract

A growing body of evidence indicates that plants can influence the survival and reproduction of the insect herbivores they host via both herbivore density-dependent and density-independent processes. A remaining challenge is identifying how density-dependent and density-independent processes in herbivores contribute to the distribution of herbivores in natural populations. I tested which herbivore recruitment parameters-the intrinsic rate of increase, carrying capacity, or shape of density dependence-contributed to variance in the distribution of a gall-making fly among individuals of its host plant by experimentally manipulating herbivore density on plants in the field. I used model selection to determine the relationships between herbivore demographic parameters and the natural, pre-experimental pattern of herbivore abundances. The naturally occurring pattern of herbivore abundances before the experiment covaried positively with the herbivore carrying capacity, a parameter inversely related to the strength of density dependence, but not with the shape of density dependence or the intrinsic reproductive rate. This means that plants with high natural herbivore abundances had lower herbivore density dependence but not higher rates of herbivore reproduction at low abundances. More generally, these results suggest that density dependence mediated through the host plant was responsible for the significant spatial variance in abundance of this herbivore among host plants. This also means that the processes influencing the spatial variance in the abundance of this herbivore occur at high, but not at low, herbivore density. This suggests that when measuring parameters of herbivore preference for, or performance on, plants with different genotypes or phenotypes, ecologists should use a range of herbivore densities to ensure that they capture density-dependent processes. Densitydependent recruitment at the scale of host plants could be a widespread determinant of abundance patterns for the many insect herbivores that have high heterogeneity in abundances among host plants and low variance in that pattern through time.

Host selection by an insect herbivore with spatially variable density dependence

“…In 2011, gall abundances on 355 plants at this study site ranged from 0 to >100: about 50 % of plants had from zero to two galls, while less than 5 % of plants had >30 galls (Wetzel 2014). The interplant variation in gall densities is related to variation in the strength of density dependence in larval survival among plants.…”

“…In our study species, Eutreta diana (Tephritidae), per capita larval survival declines strongly with increasing female density on the majority of host-plant individuals in a population and weakly on a small minority (Wetzel 2014). We expected females to seek plants with low competitor density because larval density dependence, averaged across the landscape, is strong.…”

“…It varies in form and strength among species (Sibly et al 2005) as well as among sites within species. Recent work shows that the parameters of density dependence experienced by insect herbivores can vary substantially among plant individuals within a population because of differences in a variety of nutritive and defensive traits (Agrawal 2004;Underwood 2007;Miller 2007;Wetzel 2014). These findings suggest that ovipositing females should not only respond to the density of competitors but also to traits predictive of the strength of density dependence mediated by plants.…”

“…We expected females to seek plants with low competitor density because larval density dependence, averaged across the landscape, is strong. In contrast, we expected females to ignore competitor density and to respond mainly to plant traits associated with strength of larval density dependence because competitor density has only a weak effect on fitness on those rare plants with weak density dependence (Wetzel 2014). If oviposition choices were mainly a function of the density of competitors, this would suggest that oviposition behavior were driven by the average plant, which has strong density dependence.…”

“…The many plants with low gall densities have strong density dependence, whereas the rare plants with high E. diana densities have weak density dependence. Differences in gall densities among plants are not related to densityindependent demographic parameters like the intrinsic rate of increase or fitness at low herbivore density (Wetzel 2014). The relative abundances of galls among plants are highly correlated through time (Pearson's r = 0.7-0.8) at least over 4 years, suggesting that density dependence within plants is also consistent through time (Wetzel 2014).…”