Interaction of pollinators and herbivores on plant fitness suggests a pathway for correlated evolution of mutualism- and antagonism-related traits

Abstract

Different kinds of plant-animal interactions are ordinarily studied in isolation, yet considering the combined fitness effects of mutualistic and antagonistic interactions is essential to understanding plant character evolution. Functional, structural, or phylogenetic associations between attractive and defensive traits may be nonadaptive or result from correlational selection on sets of herbivory- and pollination-linked traits. Nonadditivity of fitness effects of mutualists and antagonists, a requisite for correlational selection, was experimentally tested in the field. We created experimental populations of the insect-pollinated perennial herb, Helleborus foetidus, at 16 different locations distributed among three regions in the Iberian Peninsula. Plants experienced one of four possible selective regimes generated by independently weakening the effects of pollinators and herbivores (flower and fruit predators) according to a two-way fully factorial design. Effects were assessed in terms of number of next-generation offspring recruited per mother plant under natural field conditions. Differences among H. foetidus plants in the strength of their interactions with pollinators and herbivores translated into differential fitness, as measured in terms of recruited offspring, and subsequent changes in plant population densities. A strong, geographically consistent nonadditivity in the fitness consequences of pollinators and herbivores was found also. Plants possessing the particular combination of "traits" simultaneously enhancing pollination and escape from herbivores enjoyed a disproportionate fitness advantage over plants possessing any of the other three possible "trait" combinations. Results suggest a simple, possibly widespread ecological pathway favoring the adaptive correlated evolution of mutualism- and antagonism-related plant traits in pollinator-dependent plants suffering intense flower and fruit herbivory.

Fig. 1.

Map of the Iberian Peninsula showing the location of the three study regions. The figures in brackets indicate the number of separate H. foetidus populations (treated as blocks in statistical analyses) used in each region. In Caurel, H. foetidus plants grew at 950- to 1,350-m elevations in pine (Pinus sylvestris) plantations, Castanea sativa groves, open successional scrublands, and Brachypodium rupestre meadows. In Cazorla and Mágina, plants were in pine (Pinus nigra)- or oak (Quercus rotundifolia)-dominated forests at 900- to 1,600-m elevations.

Fig. 2.

Interaction graphs for the effects of Pollinators and Herbivores on the mean number of emerged (A) and recruited (B) seedlings of H. foetidus per experimental plot in 2001. Circles denote model-adjusted cell means with herbivores present (open circles, dashed line) or excluded (filled circles, continuous line), and vertical segments extend over ±1 SE. The values plotted are back-transforms of the results obtained from the analyses of log-transformed data (hence the asymmetry of vertical segments with respect to the mean). In each graph, P values correspond to tests of significance of the four simple main effects involved in each interaction (see Table 1 for significance tests of main effects and interactions).

Fig. 3.

Interaction graphs for the effects of Pollinators and Herbivores on the mean number of recruited seedlings of H. foetidus per experimental plot in 2001 for the three study regions (Fig. 1). Circles denote model-adjusted cell means with herbivores present (open circles, dashed line) or excluded (filled circles, continuous line), and vertical segments extend over ±1 SE. Note the differences between regions in vertical scale.