Testing Hypothesized Latitudinal Patterns in Plant-insect Interactions from Tropical to North-temperate Regions

摘要

Biotic interactions vary geographically, with potential consequences for evolutionary processes such as local adaptation and speciation. The "biotic interactions hypothesis" posits that biotic interactions are stronger at lower latitudes, and that coevolution has contributed to the well-known, poorly understood pattern of greater biodiversity toward the equator. The hypothesis plausibly connects known environmental variation to speciation rates, but strong empirical support is lacking for latitudinal patterns of some interactions, particularly herbivory and pollination. Current data may be insufficient, and methods of quantifying interaction strength lack standardization and rigor. I tested the hypothesis that plant-insect interactions are stronger at lower latitudes by quantifying contemporary herbivory and pollination rates, and patterns of trait evolution that may show signatures of past selection. Herbivory and plant defense were measured along a 15° latitudinal gradient in Phytolacca americana in the eastern USA, and in P. rivinoides, its tropical congener, in Costa Rica. Both are perennial early-successional herbs studied in their native ranges. Patterns of herbivory rates in the field and plant defense (quantified using a bioassay in a common garden) were consistent with the biotic interactions hypothesis for most metrics, particularly for young leaves, which are rarely measured in comparable studies. Variation in leaf chemistry, which is a strong predictor of palatability for a generalist herbivore, depends strongly on geography and leaf age. In addition, multiple metrics of pollination (e.g., self-pollination ability) were evaluated for P. americana at the northern and southern range edges, and for P. rivinoides in Costa Rica. Patterns were consistent with a greater reliance on insect pollination at lower latitudes. For both herbivory and pollination, latitudinal patterns were rarely clinal, instead demonstrating thresholds or scale-dependency. For example, comparisons of P. rivinoides and north-temperate P. americana were consistent with the biotic interactions hypothesis, but comparisons of either region with subtropical P. americana showed a variety of patterns. These three chapters strongly support a long-standing hypothesis, and generate new hypotheses for ecological mechanisms driving geographic variation in biotic interactions. The fourth chapter reviews the state of evidence for evolution and genetics of mutualisms, focusing on the question of how adaptation and speciation may be unique when driven by mutualistic interactions. Together, these studies further our understanding of the ecological causes and evolutionary consequences of variation in plant-insect interactions.