Image of Tularosa Creek valley.

Research

My primary research focus concerns the biogeography and conservation of mammals. My students and I form the Mammalian Diversity and Conservation Laboratory. The common question that unites this work is: "what lives where and why" I address this question both at the very broad spatial and temporal scale of ecological and historical biogeography (i.e., insular systems and continental regions) as well as at the finer scale of locality-specific habitat. As a mammalogist, my research is focused on mammals, especially non-game, small game, and sensitive species. Data for my research is generated through both field studies and from museum research collections. Answers to the unifying question contribute to a basic understanding of mammals, and they form the basis for understanding many conservation and management problems, such as extinction, biological invasions, and response to environmental change. Thus, this research typically has direct conservation and management applications. Much of it can be categorized within the new and growing discipline, Conservation Biogeography.

Insular Biogeography

Mountaintop Coniferous Forest Islands.-I study mammalian faunas on insular systems as models for understanding processes, such as dispersal and extinction, that shape species distribution patterns and community structure. Insular systems include true islands surrounded by water, as well as patches of habitat within a matrix of dissimilar habitat. Much of my research has focused on islands of coniferous forest on mountaintops in western North America (including west Texas) that are isolated by seas of arid low elevation desert and grasslands. For example, previous work by J. Brown and M. Lomolino on montane mammal biogeography in the American Southwest has become a paradigm about mammal distribution patterns and how they have been structured as a result of past climatic shifts. They contend that modern mammal communities are structured through the processes of local extinction as well as intermontane dispersal. In contrast, we (with T. Yates and M. Bogan) came to dramatically different conclusions (in press in Ecography). Our approach was novel through use of techniques from the divergent disciplines of ecological and historical biogeography. Species data were based on my field surveys and museum records. Importantly, we were able to include the variable of island age, which was generated through a novel cladistic-like analysis of valley barrier features. This was the first time that island age has been included as a variable in any similar study on continental systems and one of the few for any island system. We found that island age and area were the only determinants of species diversity patterns. Hence, extinction was the primary ecological process shaping distribution patterns. Currently, I am expanding on this research by examining nestedness in this system. This allows for the identification of specific montane populations that are most at risk of extinction, and hence will contribute to conservation planning. As a further expansion of this research, I am currently writing an NSF grant to extend my island age methodology to montane islands in the Great Basin, which has been the center of controversy surrounding interpretation of species richness patterns in montane systems.

As a second example of my biogeographic research on montane systems, I highlight a collaborative research project (with A. Ditto) that is in press in Journal of Biogeography. Evolutionary theory makes predictions about patterns of genetic variation on insular systems, and these patterns have been well established through empirical evidence on oceanic islands: genetic variation exhibits a positive correlation with island area and a negative correlation with island isolation. We tested these patterns for a suite of montane mammals in the American Southwest and found that the patterns of genetic variation were consistent with evolutionary predictions. We then developed 3 different statistical models to predict levels of genetic variation expected under different climate warming scenarios that would result in a loss of area of coniferous forest on mountaintops. We found that most populations were predicted to experience some loss of genetic diversity and that, depending on the dimension of genetic diversity involved (e.g., allelic diversity, polymorphism, heterozygosity), these changes would put populations at heightened risk of extinction in both the short and long-term. As an extension of this research, I am currently preparing a manuscript that details patterns of genetic variation in another species, the montane vole (Microtus montanus). In this species, patterns of genetic variation were not consistent with expectations of evolutionary theory (e.g., the smallest, most isolated population had the highest genetic variation). I propose a climate-induced metapopulation model that accounts for this pattern.

Texas Coastal Islands.-Species diversity patterns often behave in different and unexpected ways on very small islands. Unlike most insular systems, which exhibit a significant positive relationship between island area and species richness, there is often no relationship between island area and species richness on small islands (i.e., the graphical relationship is a horizontal line). This pattern is known as the Small Island Effect (SIE), but it has been rarely investigated and processes contributing to these patterns are virtually unknown. I am investigating these patterns with one of my graduate students (G. Jones). This research is focused on a series of very small (0.05-6.9 ha) islands located in the Laguna Madre off the coast of Texas. Some of these islands were human-created from dredged material during construction and maintenance of the Gulf Intracoastal Waterway, while others were of natural origin. We sampled terrestrial mammals and physical and habitat characteristics of these islands in order to examine patterns of species diversity. SIE is typically detected using simple break-point regression that identifies the inflection point between the horizontal line found for small islands (i.e., the SIE) and the positive slope found for larger islands. However, we used a novel method that allows for the identification of multiple breakpoints, including stair-step patterns. In addition, in order to put these islands into context of the larger insular system of the Gulf of Mexico, we also included analyses of mammal faunas on other larger islands based primarily on museum records and our own field surveys on Padre Island. We found that breakpoint models performed better than linear models in most situations, which supported the SIE. Further, most breakpoint models were simple (i.e., a horizontal line for small islands followed by a positive slope for larger islands). However, on small islands, inclusion of both small and large mammals resulted in a stair-step pattern, which suggests that there are different processes determining distribution patterns for resident small mammals as compared to non-resident large mammals. Our research on this system is continuing, especially by incorporating habitat data and potential for competition into the model.

Habitat Relations and Response to Human Land-Use

I am interested in research on species habitat relationships, especially in wetland, grassland, and coniferous forest ecosystems. At the most basic level, this research involves habitat analyses of individual species in order to better understand their biology. Examples of these include least shrew (Cryptotis parva), Abert's squirrel (Sciurus aberti) and red squirrel (Tamiasciurus hudsonius). I am also interested in habitat partitioning between interacting pairs of sympatric species. This research can provide unique perspective into ecological factors associated with species distribution patterns, and in the case of rare or sensitive species, can provide important insight into appropriate habitat management. For example, with one of my graduate students (J. Malaney) we investigated habitat partitioning between snowshoe hare (Lepus americanus) and mountain cottontail (Sylvilagus nuttallii) in a zone of sympatry in northern New Mexico. We found that while snowshoe hare were essentially restricted to dense spruce-fir stands, mountain cottontail were more cosmopolitan and occupied virtually all available habitats, especially those that had less dense cover. We concluded that traditional methods for managing snowshoe hare (i.e., patch clearcuts) that were developed in regions that lacked another sympatric leporid were inappropriate in New Mexico because they were likely to favor mountain cottontail. In another example comparing the sympatric meadow jumping mouse (Z. h. luteus) and western jumping mouse (Z. princeps) in northern New Mexico, I found significant differences in habitat associations (e.g., Z. h. luteus is a specialist of emergent wetlands, while Z. princeps is more of a generalist), which has called for a major reinterpretation about the ecology and biogeography of the endangered Z. h. luteus.

I am especially interested in the response of mammals to human land-use practices. These studies typically have direct management application. For example, one of my students (C. Wampler) recently defended her thesis that evaluated the effects of fuel reduction thinning treatments on terrestrial mammals in a mixed coniferous forest ecosystem. We found that thinning treatments appeared to benefit small mammals, carnivores, and ungulates, in comparison with older control stands that had not been cut in ca 60-100 years. However, there was little difference among the 3 types of thinning treatments and they did not differ from younger control stands that had not been cut in ca 20-30 years. Thus, this research demonstrated that more labor intensive and economically expensive methods did not provide an added benefit to wildlife and that thinning provided no added benefit to wildlife if applied to younger stands.

I have two ongoing research studies that involve habitat relationships and temporal changes in biogeographic patterns in Chihuahuan desert grassland mammals. Due primarily to livestock grazing and subsequent processes of desertification, there has been a dramatic decline in these grasslands and their conservation has been recognized as a priority. In a large collaborative project (with G. Roemer, E. Geffen, M. Moses) we are studying functional demographic relationships in the banner-tailed kangaroo rat (Dipodomys spectabilis), which is an indicator of well-developed desert grassland. Once the functional models have been developed based on our long-term mark-recapture plots and associated environmental data, we will develop a spatially explicit metapopulation model that will serve as a means for conservation planning. In a second study, I am examining temporal shifts in the distribution of two grassland species, the deer mouse (Peromyscus maniculatus blandus) and the white-footed mouse (Peromyscus leucopus), in the Chihuahuan Desert region. This is a synthetic study that is drawing on multiple sources of data including historical museum records, previous behavioral and habitat studies, and current long-term small mammal studies. Preliminary results indicate that P. maniculatus was the dominant species in the region during the early 1900s, while P. leucopus had an extremely restricted distribution. However, the situation has now reversed with P. leucopus being the more common and widespread species. One primary reasons for these changes is the loss of open grasslands through shrub invasion, which is favored by P. leucopus.

Inventory and Monitoring

I have been very successful obtaining funding from state and federal agencies for inventory and monitoring studies. The scope of these studies is typically expanded to include analyses on other aspects of a species' biology or community structure. My inventory studies include status assessments for endangered species, such as least shrew (Cryptotis parva), least chipmunk (Tamias minimus atristriatus), montane vole (Microtus montanus arizonensis), and meadow jumping mouse (Zapus hudsonius luteus). Such studies have been important to both the state and federal governments in assessing appropriate legal status for these species. For example, based on my extensive research on Z. h. luteus, this species was recently elevated to endangered in New Mexico and it is under review for potential candidacy listing on the ESA. My inventory studies also include comprehensive mammal surveys of entire areas, such as Padre Island National Seashore, Sugarite Canyon State Park, and Grand Canyon National Park. My monitoring studies have generally been focused on specific US Forest Service management indicator species, including Abert's squirrel (Sciurus aberti) and red squirrel (Tamiasciurus hudsonicus). However, I was also part of collaborative research team that recently finished a 3-year study to develop a monitoring plan for terrestrial resources along the Colorado River in the Grand Canyon. This research, particularly the inventory studies, requires close association with museum research collections. Species are documented through the collection of specimens and existing specimen records are used to reconstruct historical distribution patterns.

Species Ecology

As a mammalogist, I am keenly interested in the ecology of individual species of mammals. Most of my research in this area has focused on distribution, habitats, and other natural history aspects of species that are either rare or poorly known. Examples of some of the species that I have recently been studying include nine- banded armadillo (Dasypus novemcinctus), Arizona gray squirrel (Sciurus arizonensis), western heather vole (Phenacomys intermedius), and ermine (Mustela erminea). In most cases, I am interested in the ecology of species at their range limits. The ecology of an organism is often different at the range margin and this can shed unique light on such things as habitat requirements, required species interactions, and factors controlling demography. For example, in collaboration with one of my graduate students (J. Malaney) we studied the demography and other natural history characteristics of snowshoe hare (Lepus americanus) at its extreme southern range limit in northern New Mexico. We found a significant latitudinal gradient in various demographic characteristics, including body condition, density, and home range size, which were all lower in New Mexico. These results supported a hypothesis that habitat fragmentation was the major determinant of population demographics of southern snowshoe hare populations.

Systematics and Taxonomy

My graduate work was in the area of systematics and historical biogeography. My dissertation focused on the assertion that our understanding of the relationships among the concepts of historical ecology, biogeography and evolution were incomplete and involved the theoretical synthesis of these ideas including: 1) the reintroduction of the centrifugal speciation model into the framework of recent phylogenetic thinking; 2) evaluation of the interrelationships between the "history of place" and the "history of lineage", and 3) an accounting of the relationships among anagenesis, cladogenesis, and particular modes of speciation in light of various ecological factors. I proposed a hypothetico-deductive method for testing among alternative models of speciation, which relied on historical and contemporary distributional patterns of organisms along with specific phylogenetic and biogeographic predictions of alternative modes of speciation. I then demonstrated that the theoretical research program could be carried out with actual data by providing an empirical example based on phylogenetic analyses of southwestern montane mammals.

Currently, I am involved with collaborative research projects focused on the systematics and taxonomy of several species of small mammals in the American Southwest. For example, one of my graduate students (A. Montoya) is examining morphological variation in the gray-footed chipmunk (Tamias canipes), which is endemic to the mountains of southeastern New Mexico and extreme western Texas. We refined patterns of geographic variation in this species, which will result in a new subspecies arrangement. I am also collaborating (with D. Crawford and D. Moore) on the phylogeography of the Mexican vole (Microtus mexicanus) species group based on diverse data sets (morphology, allozymes, chromosomes, cytochrome-b). Finally, I am collaborating with the Museum of Southwestern Biology (J. Cook, J. Malaney) on a morphological and cytochrome B sequence analysis of jumping mice (Zapus) in the American Southwest. This project is an extension of my larger research on the ecology and conservation status of the meadow jumping mouse (Z. hudsonius luteus).

Use and Interpretation of Museum Specimen Data

I am interested in research concerning the use and interpretation of data derived from museum specimen records. Issues with regards to interpretation of descriptive locality data are particularly important since these data are often used in GIS modeling. For example, I am in the initial stages of a research project aimed at evaluating changes in GIS habitat models based on a comparison of naive versus expert-interpreted georeferencing of descriptive locality data. Preliminary investigations suggest dramatically different interpretation about habitat, especially for specialist species.

Another aspect of my research concerns the interpretation of species distributions when there is a paucity of credible occurrence data, such as museum specimen records. My recent paper in Biological Conservation reviewed this topic, which included a method for such interpretation. More specifically, I provided an overview about potential problems that can occur when interpreting museum data and provided a structure for evaluating the credibility of different types of occurrence data. I then recommended a method for inferring a species distribution, which was based on evaluation of existing occurrence records, assessing reasons for the paucity of records, evaluating habitat connectivity, and evaluating biogeographic patterns of other species. I then provided an example of the method by considering the potential distribution of Canada lynx (Lynx canadensis) and wolverine (Gulo gulo) in New Mexico.

Image of Tularosa Creek valley