Predispersal seed predators and nectar robbers: The influence of plant and animal traits on plant reproduction and bumblebee foraging behavior

The field of ecology is devoted to the study of patterns and processes that govern species distribution and abundance. This thesis focuses on the interactions (antagonistic and mutualistic) between plants and insects to advance our understanding of 1) how predispersal seed predators may influence the reproductive fitness of narrowly distributed versus widespread plant species and 2) the ecological effects of an introduced bumblebee, Bombus terrestris L., for a native congener, Bombus dahlbomii Guerin-Meneville, and a native shrub, Fuchsia magellanica Lam. in southern Chile. Conservation and management issues related to rare species biology and invasion biology are discussed in the context of my research findings.;In Chapter I , I have two main research objectives; I) I test the hypothesis that predispersal seed predation rates are higher for a narrow endemic species, Astragalus sinuatus Piper, compared to its sympatric and widespread congeners, Astragales leibergii Jones and Astragales purshii Dougl. ex. Hook. var. purshii and 2) I examine plant and insect traits that may be responsible for observed patterns. I found that seed loss due to predispersal seed predators was significantly higher and seed production was significantly lower for the rare species than common congeners of two years of study. Further, trait comparisons among the rare and common species suggest that high seed predation rates in the rare species reflect a combination of factors: plant phenology, seed predator identity and phonology, plant vigor, and plant dispersal ability. The results support the hypothesis that rare species may be more susceptible to predispersal seed predation relative to common congeners and may be an important contributing factor of plant rarity. In addition, rare plants may possess certain traits that predispose them to high predation rates and/or narrow distributions.;In Chapter 2, 1 have two main research objectives; I) I further test the 'rarity-predation' hypothesis but on larger geographic scale, using five sympatric, congeneric rare- common pairs and 2) I explore how- plant traits and insect identity may influence predation rates. I use rare-common species in the genus Astragalus L. as a model system to test this hypothesis. The comparative approach used in Chapter 1 and 2 served to eliminate and/or reduce evolutionary, environmental and site-specific biological factors that confound interpretation of predation patterns between rare-common species. My results indicate that some rare Astragalus species are more susceptible than sympatric congeners to seed predation, but others are not. Thus, in general, the 'rarity-predation' hypothesis was not supported. For rare plants that were more susceptible to seed predation, traits such as larger seed mass, greater pericarp width, greater planar area, and higher per capita flower production may influence host-plant selection. Insect identity and feeding mode played a large role in the magnitude of seed loss suggesting the need to better understand the identity and ecology of seed predators to help explain patterns. Further, regardless of rare-common designation, higher predation rates occurred among species with larger seeds, pods with a greater pericarp width and, in some cases species that produced more flowers. The study of both intrinsic factors such as plant traits and extrinsic factors such as predispersal seed predation were important to advance our understanding of plant rarity.;In Chapter 3, I investigate potential ecological consequences of an introduced nectar robbing bumblebee, B. terrestris, for a native congener, B. dahlbomii and a native shrub, F. magellanica in Chile. My research objectives include 1) investigating foraging behavior and abundance of B. terrestris and B. dahlbomii in two sites where they co-occur and two sites where only B. dahlbomii is found, 2) quantifying B. terrestris robbery rates on F. magellanica and determine if F. magellanica is pollen limited and 3) examining bumblebee and floral traits that may help explain robbery patterns. My results show that relative abundance of B. terrestris was 11-19 times more abundant than B. dahlbomii where they co-occur. Bombus dahlbomii was ∼9-18 times more abundant when B. terrestris was absent. Visitation frequency did not differ between B. terrestris and B. dahlbomii but visitation frequency of B. dahlbomii was significantly lower at sites where B. terrestris was absent. Floral robbery rates in F. magellanica ranged from 83-97%. The pollen limitation experiment showed fruit and seed set were significantly greater with pollen addition, indicating that legitimate pollinators may be important for reproductive success in F. rnagellanica . Finally, plant-insect trait comparisons suggest that nectar robbing by B. terrestris is likely due to its short tongue length relative to the depth of the corolla in F. magellanica.