Human activities have drastically altered Earth's ecosystems. Chief among these changes is the alteration of habitats and the conversion of forests, pastures, and rangelands into homes, schools, and commercial and industrial sites. Housing development (hereafter referred to as urbanization) in particular is spreading rapidly. The ecological consequences of urbanization have received considerable attention, but what are the evolutionary implications? Striking examples have shown that direct human actions, such as overharvesting, hunting, and pest management, can affect patterns and rates of evolution in the species that are the targets of these activities. Yet, the degree to which urbanization alters the direction and strength of natural selection on native, non-target species, especially plants, remains largely unexplored. The goal of this study is to understand how urbanization alters patterns of natural selection on a native plant through changes in species interactions, including pollination, floral herbivory, and nectar robbing. The proposed work provides a powerful approach to document changes in the patterns of species interactions due to urbanization, isolate interactions responsible for changes in natural selection using experimental studies, and predict and test mechanisms responsible for such changes. Collaboration with the North Carolina Botanical Gardens will facilitate dissemination of research results to the public, and the principal investigators will present one lecture per year to the local garden club. The project also provides training opportunities to undergraduate and graduate students. Participation in research can be a critical learning experience in the education of many students, and this proposal provides for human resource development at multiple academic levels.
Species evolve in response to their environment. While the study of how species evolve in natural contexts has been studied for more than a century, only recently are biologists recognizing that species also evolve to adapt to human-constructed environments. Because the fastest form of land change on the planet is the conversion of forests and fields to urban/suburban developments, it is important for biologists to understand how such changes will affect the evolution of native species. In the last decade there has been increasing research on how urbanization affects species abundance, diversity, and interactions, but much less attention has been paid to understanding how native species will evolve in response to these changes. The goal of this study was to understand how urbanization alters patterns of natural selection on native plants through changes in species interactions. This work uses observations and experiments in the field to evaluate the degree to which urbanization affects natural selection on floral and chemical defense traits in the native plant Carolina jessamine through changes in floral interactions. Our sites were located in Raleigh-Durham and Wade counties, NC, where this plant’s main floral interactions are with pollinating bees and with animals that damage flowers, including caterpillars that eat flowers and carpenter bees that pierce holes in flowers to take nectar without pollinating. We also took the first steps towards understanding what aspects of (sub)urbanization caused these changes in species interactions and plant reproduction. We collected many large data sets, including measuring floral traits, chemical defenses, floral damage, pollination, and reproduction for thousands of plants across many pairs of sites over a three-year period. Each site pair consisted of a forested tract of at least 10 acres (‘natural forest’) near a similar forest tract within single-family suburban developments (‘suburban forest’). For each site pair we also measured a range of factors that could be influenced by urbanization and affect species interactions, such as housing density, tree canopy cover, parcel size, and counts of flowers on garden plants in front yards. We also measured bee abundance, diversity, and pollinator service to three native plant species across several site pairs. Finally, we measured floral interactions and plant reproduction in 32 suburban sites chosen to vary in housing value and parcel size, factors that preliminary analyses suggested influenced interactions. Finally, we conducted a greenhouse study to determine whether floral and chemical defense traits have a genetic basis, which is necessary for traits to evolve. These measurements have created large data sets that we will make available to the public upon publication of our work, which may be useful for regional planners, urban gardeners, and conservation biologists. Much of our data is still being analyzed due to the complexity of the data sets. However, we have found consistent differences in floral damage across site pairs and years; floral damage is higher in suburban than natural forests. Natural selection on floral and chemical defense traits varied across sites and years, but to date no clear pattern has emerged indicating consistently different selection in suburban compared to forested sites. We found that floral and chemical defense traits have a genetic basis, and that floral chemical defenses are positively related to leaf defense levels. Preliminary analyses suggest that housing values are associated with floral abundance in front yards, and that floral damage is greater in suburban neighborhoods with greater tree canopy cover. Across two years, we found that there were more bees (abundance) in suburban than natural sites, but not more different kinds of bees (species richness). Bee abundance and species richness were both positively related to the abundance and richness of flowering species within forests, and the proportion of surrounding developed open areas, such as yards and roadsides, was a strong positive predictor of both bee abundance and richness. These results suggest that open habitats and the availability of floral resources in suburban sites can support abundant and diverse bee communities and underscore the potential for native bee conservation in urban habitats. Furthermore, we found that three native flowering plant species were pollen-limited for reproduction, meaning that plants would benefit from more pollination. However, two of the three species were less effectively pollinated in suburban compared to natural forests. These results raise an important policy implication, namely that increased bee abundance may not translate into increased pollination services to native plants growing in suburban remnants. Thus, managing for pollinator abundance alone may not conserve their important ecosystem services for wild-growing plants. This research supported significant student training. We involved over 40 undergraduates and recent graduates, and many of these students are now in graduate school or pursuing science careers. This research also provided partial funding for three graduate students, and created numerous opportunities to interact with the public about ecology happening in their back yards.
