<p> Biodiversity has decreased due to anthropogenic activities, and extinction rates are currently one hundred to one thousand times greater than the background rate. While the connection between deforestation and biodiversity loss is well documented within tropical rainforest ecosystems, comparatively little is known about the effects of desertification on biodiversity in dryland ecosystems. Drylands, which cover nearly half the terrestrial surface and are highly vulnerable to desertification, are among the most endangered ecosystems. To understand how biodiversity responds to environmental degradation in these fragile ecosystems, I studied arthropod diversity within a human-modified landscape suffering from deforestation and desertification in the Sahel of western Senegal. My specific objective was to determine whether arthropod, beetle, spider and ant diversity differed between protected areas of tropical dry forest and surrounding communal lands suffering from desertification. I established 12 quadrats spaced homogenously throughout each protected area as well as adjacent communal land at three different locations (Beersheba, Bandia and Ngazobil). Within each quadrat, I measured canopy closure, characterized vegetation and collected arthropods using pitfall traps during the 2014 dry (May) and rainy (September) seasons.</p><p> I collected 123,705 arthropods representing 733 morphospecies, 10,849 beetles representing 216 morphospecies, 4,969 spiders representing 91 morphospecies and 59,183 ants representing 45 morphospecies. Results showed protected areas contained greater arthropod and spider diversity than communal lands, beetle diversity varied depending on location and season, and communal lands contained greater ant diversity than protected areas (<i>P</i> &le; 0.05). My results illustrate the importance of a multi-taxa approach in understanding biodiversity response to anthropogenic disturbances. Conserving arthropod diversity in the Sahel will require the creation and preservation of more protected areas of a variety of sizes and successional stages as well as the adoption and extension of land-restorative techniques such as Zai and farmer managed natural regeneration (FMNR).</p>

<p>Bees are essential for crop pollination services; however, the ecosystem services they provide by pollinating native plants are crucial for maintaining biodiversity in natural systems. This study analyzed bee community changes within, on the edge, and outside of forest fragments. Phenological variation over two seasons and modeling to predict bee abundance associated with forest characteristics was performed. Finally the genetic diversity of two species of Bombus was analyzed. Bowl traps and aerial sweep netting in and near forest patches were used to sample bee communities in northern Delaware between March and August, 2013 and 2014. Bee phenology stayed consistent between field seasons, but fewer bees were collected in 2014 compared to 2013. Abundance modeling (R: unmarked) predicted that bee abundance was negatively impacted primarily by agriculture surrounding the forest patch. Analyses using microsatellites revealed that B. impatiens lacked overt population genetic structure while B. bimaculatus showed two genetically distinct populations. Pesticide analysis revealed 17 pesticides observed on adult bees from seven urban and suburban forest patches. Further studies investigating natural bee fluctuations in response to multiple field seasons, floral diversity, and chronic pesticide exposure are crucial in order to assess native pollinator health.

The European sheet-web spider Linyphia triangularis (Araneae: Linyphiidae) has recently become established at Acadia National Park and other areas of coastal Maine, where it reaches very high densities. Surveys indicate that three large native sheet-web spiders, Pityohyphantes sp., Frontinella communis, and Neriene radiata, are relatively scarce in both forest and coastal habitat at Acadia. Anecdotal evidence suggests that these species were more abundant at Acadia prior to the invasion, and that they are currently more abundant elsewhere in Maine (i.e., Cobscook Bay State Park) where L. triangularis populations are smaller. These native species vary in morphology and web structure; in general, Pityohyphantes sp. is most similar to L. triangularis, whereas F. communis is least similar. Despite their similar size as adults, the annual phenology of natives differs considerably from that of L. triangularis, such that natives and the invader are rarely of similar size classes. This dissertation documents the first experimental studies of interactions between L. triangularis and native spiders in Maine. Experiments conducted in the field and laboratory demonstrate negative interactions between L. triangularis and native linyphiids. Competition for habitat and aggressive web invasion appear to be more important than food competition in driving these interactions. Densities of L. triangularis comparable to those found at Acadia prevent released Pityohyphantes sp. from establishing webs on experimental plots (Chapter 1). I find no evidence that L. triangularis substantially reduces prey availability, or that natives (Pityohyphantes sp. and F. communis) remain at web sites longer if given additional prey (Chapter 2). Experiments conducted in the field (Chapter 3) and lab (Chapter 4) demonstrate that web invasion by L. triangularis is likely an important mechanism of its success at Acadia. Linyphia triangularis are more likely to usurp the webs of native spiders than vice-versa, and are also more willing to share webs. Phenological differences may affect native species differently (Chapter 4). Relative size predicts contest outcome between N. radiata and L. triangularis, whereas L. triangularis consistently wins contests with F. communis, independent of size. There is evidence that both species and size may predict outcomes of L. triangularis - Pityohyphantes sp. contests.

The behavioral research presented here revolves around jumping spiders (Araneae: Salticidae) belonging to the genus Phidippus. I studied the foraging behavior of P. audax. My findings suggest that foraging P. audax do not use chemical cues left by prey, while the wolf spider Pardosa milvina in the same experimental setup does respond to chemical cues. I studied movement patterns and the use of navigational beacons in the jumping spider P. clarus . I measured the degree of nest site fidelity with individually marked spiders and a grid of artificial nest tubes. I tested whether females used beacons to find their nests. I demonstrated that P. clarus are likely to need navigational skills, and are able to use beacons as a method of navigation. I examined the use of P. clarus as a biocontrol agent, and my results suggest that jumping spider predation is effective at reducing pest numbers and has a positive effect on plant performance. Lastly, I explored the role of male mate choice and size-assortative mating in P. clarus. I discovered that males and females pair assortatively for size, and adult males exhibit preferences for large females, which mature before smaller females. Female size is strongly correlated with the number of spiderlings that emerge.

Approximately 5,000 years ago, eastern hemlock ( Tsuga canadensis) experienced a sudden, rapid, range-wide decline most probably due to pest, disease, or climate change. An aphid-like defoliating insect, the hemlock woolly adelgid (Adelges tsugae), recently (1950’s) introduced to eastern North America has been spreading across the eastern United States. The adelgid attacks all size and age classes of hemlocks causing up to 95% mortality in affected stands. The potential for another range-wide hemlock decline has raised a number of concerns including the increased threat of wildfire. Altered fuel loadings in modern adelgid-affected stands and the effects of presumably similar changes in fuels and subsequently altered fire regimes following the prehistoric decline are examined. Fuels data from an adelgid-infested stand in Connecticut and an uninfested stand in Massachusetts were used to generate custom fuel models and predict fire behavior in each stand. Sediment cores were extracted from three sites in western Massachusetts and analyzed for fossil pollen and charcoal around the period of the prehistoric decline. Fossil data from two previously studied sediment cores from coastal Maine are included in the analysis. Results demonstrate a clear and highly significant increase in both fuel loadings and predicted fire behavior in the modern, adelgid-affected stand. Three of the coring sites reflect distinct, significant, short-lived increases in charcoal associated with the prehistoric decline; two do not. Results from the first three sites suggesting increased fire activity also were associated with changes in vegetation which indicate disturbance. Increased fire activity after the decline seems most pronounced in areas where fire was common before the decline. Results indicate that fire was not universally a significant factor driving post-decline succession. Research across a broader geographic area is needed to clarify the relationship between fire and hemlock following the mid-Holocene decline, but the results presented here suggest that managers of modern stands affected by the adelgid should include the possibility of intense fires as a threat to landscapes heavily affected by hemlock decline.

I studied the autecology, community ecology, and genetics of an endangered butterfly, the maritime ringlet (Coenonympha nipisiquit McDunnough), that inhabits a limited number of salt marshes in northern New Brunswick and in the Gaspé Peninsula of Québec. I studied the survival rate of first- and second-instar larvae in various microhabitats in a salt marsh at Daly Point Natural Reserve, Bathurst, New Brunswick. I found they survived significantly better in microhabitats dominated by Spartina patens (Aiton) Muhl. at an intermediate elevation. I investigated the tolerance of the maritime ringlet larvae to tidal submergence and compared their performance to a closely related taxon, the inornate ringlet (C. tullia inornata Edwards). The experiments revealed that the maritime ringlet possesses unique adaptations to tidal submergence. I examined the flight and oviposition behaviors of adult females in response to microhabitat. I found that they did not discriminate between microhabitats based upon the likelihood of larval survival as long as S. patens or other potential hosts were abundant. I explored the correlation between predator species richness and abundance with the larval survival rate in microhabitats. I found that predator abundance and species richness often responded negatively to increasing tidal flooding, suggesting that high larval mortality at high elevation sites can be caused by high predation pressure. Lastly, I investigated the possibility of genetic introgression between the maritime ringlet and inornate ringlet and reconstructed the phylogeny of the C. tullia-group taxa in North America. The genetic evidence did not support the possibility of large-scale genetic introgression and raised the taxonomic status of the maritime ringlet from a subspecies of holarctic C. tullia to a full species. The phylogenetic analyses suggested that the divergence of the maritime ringlet was much earlier than previously believed. My results will aid in protection and recovery of this endangered species.

The hemlock woolly adelgid, Adelges tsugae, is an invasive insect from Osaka, Japan threatening populations of eastern hemlock, Tsuga canadensis, and Carolina hemlock, Tsuga caroliniana, in the eastern United States. Previous studies of adelgid population dynamics and hemlock response in the early-1990s suggested that once infested, trees were likely to die within 4 to 6 years; widespread eradication of hemlock throughout the region was expected. However, the vast majority of hemlocks in western Massachusetts are still alive and many are not yet infested, despite the adelgid having been present in the state for 17 years. Many examples of individual trees in Massachusetts harboring adelgid for over ten years have been documented. In order to understand this discrepancy, we conducted detailed lifetable studies of the hemlock woolly adelgid at six locations in Connecticut and Massachusetts from 2004-2008. Adelgid density, fecundity, and survival data as well as temperature and precipitation data were examined to document and understand the apparent stability of adelgid populations in the northeast. These data were also used along with climate projection models to predict the range of suitable habitat for adelgid under two scenarios of carbon emissions through the year 2100. In an attempt to regulate adelgid populations, several species of biological control agents have been released, including Laricobius nigrinus Fender (Derodontidae), Sasajiscymnus tsugae Sasaji and McClure (Coccinellidae), and Scymnus ningshanensis Yu and Yao (Coccinellidae). Since few studies have addressed the efficacy of these beetles above the branch level, we conducted an experiment to test their effects on adelgid populations using whole-tree enclosures. These data, in addition to adelgid demography data will be useful in predicting future outbreaks, constructing simulations of adelgid growth and spread, and determining the amount of additional mortality needed from a biological control agent to stabilize adelgid populations.

<p> Anthropogenic trampling affects soil and litter-dwelling invertebrate communities in southern California parks, potentially altering community dynamics and hindering nutrient cycling in coastal oak woodland habitats. I collected meso- and macro-faunal community data from leaf litter and soil in frequently and infrequently trampled areas of six San Diego and three Orange County parks during fall 2011 and spring 2012. I recorded relative compaction, moisture percentage and litter mass then extracted eight soil monoliths at each park (25cm x 25cm x 5cm depth) and the detritus above them. I hypothesized frequently trampled invertebrate communities would show decreased abundance, richness and diversity and altered assemblages compared to infrequently trampled communities. Low impact areas showed higher richness and diversity in the soil layer in fall and leaf litter layer in spring. In the laboratory, I hypothesized Lumbricus terrestris earthworms would be more active, creating more large water-stable soil aggregates in low compaction treatments but found no effect of compaction on soil aggregates.</p>

<p> Fire is an important tool in the sustainable management of ecosystems at global and local scales. In addition to increasing biodiversity, fire has been shown to decrease exotic species invasion, promote growth of commercially and ecologically important trees, and reduce risk of wildfire. Missouri has a long history of anthropogenic and naturally induced fire aiding the establishment of oak and pine-dominated woodlands and savannas. Fire has been reintroduced through forest management in the region after a period of fire suppression to help retain oak-dominance in forests throughout Missouri. Research on the effects of fire is ample for many wildlife species and plants but virtually excludes insects including ants, especially in Missouri and most of the United States. Ants are considered ecosystem engineers for their contribution to soil turnover, aeration and chemical and structural modification and are important seed dispersers. The effects of prescribed burning on ant assemblage diversity, abundance, composition and function were examined in oak-hickory and oak-pine forests in the Missouri Ozark Highlands. Where fire was present annually for over sixty years, ant abundance, Generalized Myrmicinae, soil and litter nesters and small ants increased. Fire every four years for over sixty years resulted in higher Shannon diversity, Cryptic Species, litter nesters and small and medium ants. In addition, this treatment shared ants with both the control and annually burned plots. Control plots were dominated by Subordinate Camponotini, Cold Climate Specialists, wood nesters and medium sized ants. Ants were also assessed after just two fires over the course of ten years. Five and six years after fire ant assemblages were more affected by topographic position than by prescribed burning. Ant assemblages of burned and unburned sites were homogenous when compared using a Morisita Similiarity index. When comparing these treatments to those that have been burned for over sixty years using the Similarity index, both were more similar to the control. In summary, long term fire implementation results in more lasting changes in ant communities because habitat alteration is maintained over time. Habitat heterogeneity produces a more diverse assemblage of ants at the landscape scale and hence higher functional diversity. Finally, categorization of ant communities may simplify ant sampling so that the natural history of each species need not be known in order to assess ecological effects of ant assemblages associated with burn treatments.</p>

<p> <i>Arundo donax</i> is a rhizomatous perennial grass that invades riparian habitats and replaces native riparian vegetation. The galling wasp, <i> Tetramesa romana</i> (Hymenoptera: Eurytomidae), was selected for evaluation as a biocontrol agent of <i>A. donax</i> in 2009. However, knowledge is scarce on which factors influence host selection and oviposition in <i> T. romana.</i> Through my observations I documented distinct behavioral steps preceding emergence leading to oviposition in <i>T. romana,</i> and elicited an investigatory response in <i>T. romana</i> from plant volatiles extracted from <i>A. donax.</i> My results indicated that <i>T. romana</i> does use a chemical cue as a factor in host selection, but it is likely not the result of a species-specific constituent. Evidence from my study revealed that <i>T. romana</i> may use a physical cue in selecting a host as well. <i>T. romana</i> preferred round glass rods over flat glass slides, indicating that shape is an important factor in host acceptance.</p>