Plant size, latitude, and phylogeny explain within-population variability in herbivory.

Interactions between plants and herbivores are central in most ecosystems, but their strength is highly variable. The amount of variability within a system is thought to influence most aspects of plant-herbivore biology, from ecological stability to plant defense evolution. Our understanding of what influences variability, however, is limited by sparse data. We collected standardized surveys of herbivory for 503 plant species at 790 sites across 116° of latitude. With these data, we show that within-population variability in herbivory increases with latitude, decreases with plant size, and is phylogenetically structured. Differences in the magnitude of variability are thus central to how plant-herbivore biology varies across macroscale gradients. We argue that increased focus on interaction variability will advance understanding of patterns of life on Earth.

An ant's-eye view of an ant-plant protection mutualism.

Ant protection of extrafloral nectar (EFN)-secreting plants is a common form of mutualism found in most habitats around the world. However, very few studies have considered these mutualisms from the ant, rather than the plant, perspective. In particular, a whole-colony perspective that takes into account the spatial structure and nest arrangement of the ant colonies that visit these plants has been lacking, obscuring when and how colony-level foraging decisions might affect tending rates on individual plants. Here, we experimentally demonstrate that recruitment of Crematogaster opuntiae (Buren) ant workers to the EFN-secreting cactus Ferocactus wislizeni (Englem) is not independent between plants up to 5 m apart. Colony territories of C. opuntiae are large, covering areas of up to 5,000 m(2), and workers visit between five and 34 EFN-secreting barrel cacti within the territories. These ants are highly polydomous, with up to 20 nest entrances dispersed throughout the territory and interconnected by trail networks. Our study demonstrates that worker recruitment is not independent within large polydomous ant colonies, highlighting the importance of considering colonies rather than individual workers as the relevant study unit within ant/plant protection mutualisms.

Integrating quality and quantity of mutualistic service to contrast ant species protecting Ferocactus wislizeni.

Generalized, facultative mutualisms are often characterized by great variation in the benefits provided by different partner species. This variation may be due to differences among species in the quality and quantity of their interactions, as well as their phenology. Many plant species produce extrafloral nectar, a carbohydrate-rich resource, to attract ant species that can act as "bodyguards" against a plant's natural enemies. Here, we explore differences in the quality and quantity of protective service that ants can provide a plant by contrasting the four most common ant visitors to Ferocactus wislizeni, an extrafloral nectary-bearing cactus in southern Arizona. The four species differ in abundance when tending plants, and in the frequency at which they visit plants. By adding surrogate herbivores (Manduca sexta caterpillars) to plants, we demonstrate that all four species recruit to and attack potential herbivores. However, their per capita effectiveness in deterring herbivores (measured as the inverse of the number of workers needed to remove half of the experimentally added caterpillars) differs. Using these among-species differences in quality (per capita effectiveness) and quantity (number of workers that visit a plant and frequency of visitation), we accurately predicted the variation in fruit production among plants with different histories of ant tending. We found that plant benefits (herbivore removal and maturation of buds and fruits) typically saturated at high levels of ant protection, although plants could be "well defended" via different combinations of interaction frequency, numbers of ant workers per interaction, and per capita effects. Our study documents variation among prospective mutualists, distinguishes the components of this variation, and integrates these components into a predictive measure of protection benefit to the plant. The method we used to average saturating benefits over time could prove useful for quantifying overall service in other mutualisms.

Pollination mode in fig wasps: the predictive power of correlated traits.

The over 700 species of Ficus are thought to have co-speciated with their obligate pollinators (family Agaonidae). Some of these wasp species pollinate figs actively, while others are passive pollinators. Based on direct observations of mode of pollination in 88 species, we show that mode of pollination can confidently be predicted from fig traits only (anther-to-ovule ratio) or from wasp traits only (presence of coxal combs). The presence of pollen pockets is not a predictor of mode of pollination. Data, direct and indirect, on 142 species, demonstrate numerous cases of the loss of active pollination and suggest one or few origins of active pollination. Hence, active pollination, an impressive example of the sophisticated traits that may result from mutualistic coevolution, depends on selective forces that can be overcome in some species, allowing reversions. Despite frequent loss, active pollination remains the predominant mode of pollination in Ficus.

Self-pollination and its costs in a monoecious fig (Ficus aurea, Moraceae) in a highly seasonal subtropical environment.

The unusual floral phenology of most monoecious figs, related to their highly specialized pollination mutualism with agaonid wasps, combines pronounced dichogamy at the level of inflorescences and individuals with population-level asynchrony in flowering. This floral phenology ensures that outcrossing strongly predominates. Fig populations may thus be expected to possess deleterious recessive alleles that lead to inbreeding depression when selfing does occur. However, whether monoecious figs are self-compatible and whether selfing results in inbreeding depression have never been investigated. Using wasps as "pollination tools" and exploiting infrequent overlap in male and female phases on the same tree, we conducted controlled selfed and outcrossed pollination experiments in Ficus aurea. Our results show that this species is totally self-compatible. No negative effects of selfing could be demonstrated on syconium retention, number of vacant ovaries, seed set, or seed germination. However, wasp production had a tendency to be higher after self-pollination. While it is possible that inbreeding depression is expressed at later developmental stages, its absence at the early stages we examined is nonetheless surprising for a plant expected to be highly outcrossed. It is likely that selection pressures other than avoidance of inbreeding are responsible for the evolution and maintenance of the unusual floral phenology of figs.

On mutualists and exploiters: plant-insect coevolution in pollinating seed-parasite systems.

We investigate the coevolution of time of flowering and time of pollinator emergence in an obligate association between a plant and an insect that both pollinates and parasitizes flowers. Numerical analysis shows that the system in general evolves towards a time of flowering different from the time favoured by the abiotic environment. The equilibrium towards which the system evolves is a local fitness maximum (an ESS) with respect to mutational variation in flowering time but, for the insect, it can be a local fitness minimum at which selection on mutational variation in the time of insect emergence is disruptive. A consequence of evolutionary convergence to a fitness minimum is that pollinators having an earlier phenology can coexist with pollinators having a later phenology. Since late emerging insects are more likely to encounter and oviposit within previously pollinated flowers, their effect on the plant is more exploitative, leading them to function as cheaters within the system. Thus, in the long term, pollinators and exploiters are likely to be found in stable coexistence in pollinating seed-parasite systems.

Conditional outcomes in mutualistic interactions.

Interspecific interactions are traditionally displayed in a grid in which each interaction is placed according to its outcome (positive, negative or neutral) for each partner. However, recent field studies consistently find the costs and benefits that determine net effects to vary greatly in both space and time, inevitably causing outcomes within most interactions to vary as well. Interactions show 'conditionality' when costs and benefits, and thus outcomes, are affected in predictable ways by current ecological conditions. The full range of natural outcomes of a given association may reveal far more about its ecological and evolutionary dynamics than does the average outcome at a given place and time.