2013 REU Projects

Bero-Buell, Brianna (Arizona State University).  Mentor: Butch Brodie (University of Virginia).  
How does pregnancy modify behavior?  Mapping movement patterns and site fidelity in a viviparous reptile.  
Abstract:  Viviparity in squamate reptiles has evolved multiple times as a way for snakes and lizards to occupy a wider thermal range of habitats. Yet, viviparity comes with costs to female fitness such as decreased stamina and increased temperature demands. Behavioral modifications such as reduced locomotion can minimize risk associated with pregnancy and maximize reproductive output. I examined how pregnancy in the Eastern garter snake (Thamnophis sirtalis sirtalis) affects behavior by analyzing movement patterns in pregnant and non-pregnant females. Fluorescent pigment powder was used as a simple method to track short-term movements. Mark-recapture data from 2008 to present supplemented short term data and determined how movement changed in snakes that were pregnant in alternating years and whether reproductive state affected space use. I found no difference in movement, home orientation, or space use between reproductive states or experimentally relocated females. Short term movement in garter snakes may be consistent between females in a general sense though considerable variation exists among individuals.

Carrera, Stephanie (Swarthmore College).  Mentor: Vince Formica (Swarthmore College).  
Social habitat choice in B. cornutus.  
Abstract:  Considering habitat choice in organisms allows us to observe the importance of specific cues within habitats. Animals take into account temperature, predator presence, and competition when choosing or leaving a habitat. By choosing to consider the social environment as an area for choice we can look at the influential variables within a social environment. Using the system of B. cornutus we were able to ask population specific questions by isolating the beetles into sub populations by log and into social environments by fungal bracket. Through our observations of Bolitotherus cornutus we found differing results for both males and females of the species. While demographic variables we used to define a social space such as sex ratio, mean social partner body size, and social group size did not seem to influence neither male nor female choice we did find significance with individual-specific information. Smaller males had higher probabilities of leaving social environment that larger males did. Females who were courted at one sampling period were less likely to be seen her on the same fungal bracket during the next sampling period. The dimorphism in the traits and behaviors of this species seems to bleed into the dimorphism of their priorities in the local social environment. Future study should look at species with different mating systems and compare the important social variables those species have to the functional differences between the sexes within the species.

Cook, Phoebe (Swarthmore College).  Mentor: Vince Formica (Swarthmore College). 
Comparing selection in different networks.

Abstract:  Social network analysis can reveal important patterns at the population level. Selection at this level is a newly recognized phenomenon. In order to measure selection on network position, or do any other network analysis, a network must be constructed of some or all the individuals and interactions in the population. The few previous studies on selection on social network position have only analyzed one network, chosen based on the researcher’s knowledge of the system and assumptions about the dynamics of selection. For example, correlations between network position and male fitness have been studied in male-male networks of forked fungus beetles, but these analyses ignore any impacts that interactions with females may have on male fitness (Formica et al. 2012). The goals of this study were to compare the results of selection analysis on social networks in the same system four versions of a network: male-only and complete networks based off of direct interactions and proximity. Populations of forked fungus beetles, Bolitotherus cornutus, were monitored for most of a breeding season. Every male’s fitness was estimated and interaction networks were constructed. No selection on social network position was found in proximity-based networks, but the networks constructed using only direct interactions showed selection on strength and centrality. Of these, AICc scores showed the complete networks were slightly better than the male-only networks. The patterns of selection seem to be population-specific. A larger-scale study with replicated populations could tease apart what shapes these selection regimes. 

del Sol, Jillian (Hendrix College).  Mentor: Corlett Wood (University of Virginia). 
Effects of the larval environment on the development and adult phenotype of a horned beetle, Bolitotherus cornutus. 
Abstract:  Organisms live in complex environments and respond to singular and interactive ecological cues in ways that can alter many aspects of their phenotype. The role of the juvenile environment in development and subsequent phenotypic expression has been indicated to drive selection and behavior in species with discrete larval stages and phenotypic morphs. Elements of the larval environment have also been shown to contribute to phenotypic diversity both in naturally and sexually selected traits. We incorporated study of the larvae of sexually dimorphic forked fungus beetle, Bolitotherus cornutus (Coleoptera: Tenebrionidae), to investigate the responses of larval phenotypes to factors of the fixed larval environment and to obtain measures of fixed host quality. We found evidence for higher mortality in and lower quality of one fungal host, Fomes fomentarius, compared to the higher quality hosts, Ganoderma tsugae and Ganoderma applanatum. We also found a significant contribution of the larval phenotype to adult condition and obtained preliminary negative results on the extent of the contribution of variation in the larval environment to the hypervariability of adult horn and size traits.  Our results provide an example of how analyzing phenotypic responses of both the larval stage and adult stage variable environmental factors can contribute to a more holistic understanding of environmental effects on phenotypes, especially in organisms with discrete life stages.

Leichter, Sarah (North Carolina State University).  Mentor: Brian Sanderson (University of Virginia).  
Sex allocation differences in Silene vulgaris. 
Abstract:  The environment varies in its availability of crucial resources. To mitigate potential fitness consequences due to environmental variability organisms have dynamic patterns of resource allocation. Of particular interest is how organisms allocate resources to somatic and sexual functions. Single sex organisms can only sex allocate towards one component of reproduction, while hermaphroditic individuals can allocate towards both the male and female components of reproduction. There are different patterns of allocation in hermaphrodites; equal allocation to both components and an increase in allocation to one component at the expense of the other. This study investigated sex allocation patterns in the gynodioecious, perennial weed Silene vulgaris in varying soil concentrations of phosphorus. We hypothesized that as availability in the soil decreases there is an increase in allocation towards the male component of reproduction at the expense of the female component of reproduction. We found significant variance in number of ovules produced across populations, as well as significant differences between the sexes in the number of ovules produced within a population. However, soil concentration of phosphorus did not have a significant effect on either component of reproduction. Our results suggest that plasticity in allocation to female reproduction is the driving force to dioecy, contrary to the theory that plasticity in male reproduction motivates dioecy.

Marti, Hannah (St. Olaf College).  Mentor: Corlett Wood (University of Virginia).  
Cannibalism as a context-dependent behavior in Bolitotherus cornutus.
Abstract:  Cannibalism is a behavior with many potential costs and benefits, which are likely to be context-dependent. Two contexts might especially influence the cost-benefit tradeoff of cannibalism: social contexts (described as the presence or absence of kin) and nutrient environment contexts. This study sought to determine the likelihood of cannibalism across environments varying in social and nutrient quality contexts in Bolitotherus cornutus. Paired larvae were observed in lab trials using a comparison between sibling and non-sibling larval pairs to test the effect of social context and using B. cornutus’s three fungal host G. tsugae, G.applanatum, and, F. fomentarius, as environment treatments to test the effect of nutrient environment. While there was no effect of relatedness on the likelihood of cannibalism, larvae in G. applanatum cannibalized nearly twice as frequently as larvae in the other two fungal environments. However, patterns in cannibalism frequency did not follow the expected and widely observed pattern of highest cannibalism frequencies in the most nutrient poor environments, raising questions as to the factors that drive a differential cannibalism response in this species as well as the adaptiveness of such a difference across contexts.

Post, Alison (University of Maryland).  Mentor: Laura Galloway (University of Virginia).  
Factors influencing rates of autogamy among populations.
Abstract:  As the ice sheet over North America started to recede at the end of the Pleistocene, plant species previously restricted to southern refugia were able to expand their ranges northward.  As a consequence of migration, within these northern populations, genetic diversity decreased, homozygosity increased, and the amount of inbreeding depression decreased. Therefore, self-pollination was not as large a fitness cost, so higher rates of autogamy could evolve in these northern populations.  This pattern is found in the protandrous woodland herb, Campanula americana, commonly referred to as the American Bellflower.  It has a primarily out-crossing mating system, however, it is also completely self-compatible.  This study looked at possible mechanisms behind the increasing autogamy rates over a geographic gradient observed in C. americana.  It explored differences in the timing of floral gender phases, as well as differences in floral morphology, particularly stigmatic lobe curling, between five different C. americana populations exhibiting a range of autogamy rates.  Differences between populations were found in time to onset of female phase, however, all populations became female while pollen viability was still high. Between populations, the rate of pollen degradation and the rate of stigmatic lobe curl differed.  The populations followed 2 patterns- the same populations that had a faster rate of pollen degradation also had a faster rate of stigmatic curling, suggesting an overall accelerated floral life cycle. It was also found that greater amounts of stigma curl correlated with greater self pollen deposition onto the stigmatic lobes, signifying a possible mechanism for delayed self-pollination.  However, these differences between populations did not correspond to the latitudinal variation in autogamy rates, suggesting that a different mechanism is behind these variations in autogamy rates observed among populations.

Rosenthal, Lisa (University of California, Berkeley).  Mentors: Henry Wilbur and Becky Wilbur (University of Virginia).  
Legacy of logging on American Chestnut population dynamics and genotype decline.
Abstract:  The chestnut blight fungus, Cryphonectria parasitica (Murr.) Barr, infects the American chestnut (Castanea dentata (Marsh.) Borkh.) before it reaches sexual maturity, but leaves the root collar uninfected, allowing it to resprout after infection. Although this cycle can persist for decades, individuals are dying and those genotypes are lost forever. This study examines the legacy of logging on genotype decline and the changes in population dynamics. Davelos & Jarosz (2004) studied isolated chestnut populations similar to those when the blight first emerged in the early 20th century. We contrasted three chestnut populations of different ages in the Southern Appalachians originating from 1861, 1921 and 1985, and constructed population projection matrices, allowing us to further analyze the differences. The oldest populations had the highest disease incidence and diverged most from the other two populations. Density of chestnuts was greatest in the 1921 populations due to the size structure of the chestnuts when the blight first emerged, as well as the community structure. The Davelos & Jarosz (2004) population was declining at a slower rate than all of the populations from this study, largely due to the existence of reproductive trees. This study is intended to broaden the ecological knowledge of blight-suppressed chestnut populations, which may be of use to forest managers. 

Sacco, Laurel (University of Maine).  Mentor: Butch Brodie  (University of Virginia).  
The integration of toxicity, antipredator behavior, and color in the Red Spotted Newt (Notophthalmus viriscedens
Abstract:  Aposematic signaling is an extremely complex form a defensive strategy. It has long been known that aposematic, or warning, signals function to advertise noxious or hazardous properties to predators. The antipredator strategies of salamander species often involve chemical defense and the salamanders produce aposematic displays. In salamanders, posturing is used to display aposematic coloration. Aposematic displays in salamanders vary widely across species in both coloration and posturing, but these displays can also vary within a species. A salamander with a complex aposematic defensive system is Notophthalmus viridescens, the red spotted newt. The defensive system of this salamander is comprised of three integrated parts: an extremely potent neurotoxin, bright warning coloration and defensive posturing. The life cycle of Notophthalmus viridescens does not progress directly from the larval stage to the adult, there is an intermediate terrestrial stage known as the red eft. The red eft and adult of Notophthalmus occur in drastically different environments and, as a result, experience dramatic changes in physiology and behavior. The focus of this research is the effect of ontogenetic change on the integrated defense traits in the complex antipredator system of Notophthalmus viridescens. This study is an investigation of the correlation between toxicity, color brightness, and posture intensity within and between the red eft, intermediate, and adult stages.

Yantes, Austin (University of Wisconsin, Madison).  Mentor: Malcolm Augat (University of Virginia).
Phenotypic plasticity across varying soil environments in Silene vulgaris.
Abstract:  Environmental heterogeneity can elicit plastic responses in phenotypes as well as prompt local adaptation to arise. Sex characters are especially likely to be plastic as they are costly to produce and are often under selection pressures. Gynodioecy provides an interesting system in which to study plasticity of sex characters, as the primary sex characters contributing to fitness differs between females and hermaphrodites. The variance in sex ratios we see among wild gynodioecious populations can be partially attributed to genetics, but much of the variation is unaccounted for. The present study further explores phenotypic plasticity across differing soil environments, and the degree of local adaptation. 723 Silene vulgaris seedlings were transplanted into 5 different soils, and floral and vegetative phenotypes were collected for one month. In addition, a germination assays was performed in which seeds from 5 wild populations were planted in 6 soil treatments. There was no effect of soil treatment on sex (P = 0.425), nor did dam affect sex (P = 1.000). The effect of dam on floral phenotypes was significant, but soil treatment did not have an effect on floral phenotypes. Soil treatment had an effect on multiple vegetative phenotypes. There was also a treatment effect on whether or not an individual survived to the end of the study (P < 0.001). Germination success was affected by both seed source population (P < 0.001) and soil treatment (P < 0.001). Additionally, there was a soil by population interaction effect on germination success (P = 0.016); however, this is not explained by whether or not a seed is planted in its own soil (P = 0.839). These results indicate that the effect of the environment may vary across life stages, and that the influences of genetics and the environment on phenotypes cannot be easily disconnected. Instead, the phenotypes we see expressed in wild populations are the products of intricate and dynamic interactions between numerous genetic and environmental factors.