Brian R. Mitchell

Guidelines for a graph‐theoretic implementation of structural equation modeling

Structural equation modeling (SEM) is increasingly being chosen by researchers as a framework for gaining scientific insights from the quantitative analyses of data. New ideas and methods emerging from the study of causality, influences from the field of graphical modeling, and advances in statistics are expanding the rigor, capability, and even purpose of SEM. Guidelines for implementing the expanded capabilities of SEM are currently lacking. In this paper we describe new developments in SEM that we believe constitute a third-generation of the methodology. Most characteristic of this new approach is the generalization of the structural equation model as a causal graph. In this generalization, analyses are based on graph theoretic principles rather than analyses of matrices. Also, new devices such as metamodels and causal diagrams, as well as an increased emphasis on queries and probabilistic reasoning, are now included. Estimation under a graph theory framework permits the use of Bayesian or likelihood methods. The guidelines presented start from a declaration of the goals of the analysis. We then discuss how theory frames the modeling process, requirements for causal interpretation, model specification choices, selection of estimation method, model evaluation options, and use of queries, both to summarize retrospective results and for prospective analyses. The illustrative example presented involves monitoring data from wetlands on Mount Desert Island, home of Acadia National Park. Our presentation walks through the decision process involved in developing and evaluating models, as well as drawing inferences from the resulting prediction equations. In addition to evaluating hypotheses about the connections between human activities and biotic responses, we illustrate how the structural equation (SE) model can be queried to understand how interventions might take advantage of an environmental threshold to limit Typha invasions. The guidelines presented provide for an updated definition of the SEM process that subsumes the historical matrix approach under a graph-theory implementation. The implementation is also designed to permit complex specifications and to be compatible with various estimation methods. Finally, they are meant to foster the use of probabilistic reasoning in both retrospective and prospective considerations of the quantitative implications of the results.

Monitoring and evaluating the ecological integrity of forest ecosystems

“Ecological integrity” provides a useful framework for ecologically based monitoring and can provide valuable information for assessing ecosystem condition and management effectiveness. Building on the related concepts of biological integrity and ecological health, ecological integrity is a measure of the composition, structure, and function of an ecosystem in relation to the system’s natural or historical range of variation, as well as perturbations caused by natural or anthropogenic agents of change. We have developed a protocol to evaluate the ecological integrity of temperate zone, forested ecosystems, based on long-term monitoring data. To do so, we identified metrics of status and trend in structure, composition, and function of forests impacted by multiple agents of change. We used data, models, and the scientific literature to interpret and report integrity using “stoplight” symbology, ie “Good” (green), “Caution” (yellow), or “Significant Concern” (red). Preliminary data indicate that forested ecosystems in Acadia National Park have retained ecological integrity across a variety of metrics, but that some aspects of soil chemistry and stand structure indicate potential problems. This protocol was developed for the National Park Service Vital Signs Monitoring Program and holds promise for application in the temperate zone, forested ecosystems of eastern North America.

Coyote movements and social structure along a cryptic population genetic subdivision

A recent region‐wide study determined that the central California coyote (Canis latrans) population was genetically subdivided according to habitat bioregions, supporting the hypothesis that coyotes exhibit a dispersal bias toward their natal habitat type. Here, we further investigated this hypothesis using radio‐collared coyotes captured on a 150‐km2 study site on the border of (i.e. overlapping) two bioregions (Great Valley and Cascade Mountains). As predicted, most coyotes were assigned (based on a priori genetic criteria) to genetic clusters corresponding to one of these two bioregions. All of those assigned to the Great Valley genetic cluster were caught in (and for the most part, remained in) the Great Valley bioregion. However, contrary to expectations, the coyotes assigned to the Cascades genetic cluster occurred commonly in both bioregions. Nearly all resident individuals on the study site, regardless of the particular bioregion, were assigned to the Cascades genetic cluster, whereas a sizable fraction of nonresident (transient or dispersing) coyotes caught in the Great Valley bioregion were assigned to the Great Valley cluster. Even among resident coyotes, interrelatedness of packs was greater within than between bioregions, and packs with territories overlapping both bioregions were more closely related to those with territories completely within the Cascades bioregion than territories completely within the Great Valley bioregion. Finally, direct estimates indicated that gene flow was twice as high from the Cascades bioregion to the Great Valley bioregion than in the reverse direction. Collectively, these findings reveal the anatomy of the genetic subdivision as beginning abruptly at the bioregion boundary and ending diffusely within the Great Valley bioregion.

Coyote depredation management: current methods and research needs

Abstract This paper examines the severity of livestock depredation by coyotes (Canis latrans), reviews evidence implicating breeding (or “alpha”) coyotes in the majority of incidents, evaluates currently used depredation control techniques, and suggests directions for future research. Nonlethal control ranges from varied animal husbandry practices to coyote behavioral modification or sterilization. These methods show significant promise but have not been proven effective in controlled experiments. Therefore, many livestock producers rely on lethal control, and most employ nonselective strategies aimed at local population reduction. Sometimes this approach is effective; other times it is not. This strategy can fail because the alpha coyotes, most likely to kill livestock, are the most resistant to nonselective removal techniques. An alternative is selective lethal control. Livestock Protection Collars (LPCs) and coyote calling are the primary selective lethal approaches. However, LPCs do not have support from the general public due to the toxicant used, and the factors affecting the selectivity of coyote calling have not been studied. The greatest impediments to effective coyote depredation management currently are a scarcity of selective control methods, our lack of understanding of the details of coyote behavioral ecology relative to livestock depredation and wild prey abundance, the absence of solid research examining the effectiveness of different control techniques in a variety of habitats and at multiple predation intensities, and the dearth of rigorous controlled experiments analyzing the operational efficacy of selective removal versus population reduction.

INFORMATION CONTENT OF COYOTE BARKS AND HOWLS

ABSTRACT The information content of coyote (Canis latrans) vocalizations is poorly understood, but has important implications for understanding coyote behaviour. Coyotes probably use information present in barks or howls to recognize individuals, but the presence of individually-specific information has not been demonstrated. We found that coyote barks and howls contained individually specific characteristics: discriminant analysis correctly classified barks of five coyotes 69% of the time and howls of six coyotes 83% of the time. We also investigated the stability of vocalization characteristics at multiple distances from the source. Recordings were played back and re-recorded at 10 m, 500 m, and 1,000 m. Vocalization features were measured at each distance and analyzed to determine whether characteristics were stable. Most howl characteristics did not change with distance, and regardless of the distance discriminant analysis was 81% accurate at assigning howls among six individuals. Bark characteristics, however, were less stable and it is unlikely that barks could be used for individual recognition over long distances. The disparate results for the two vocalization types suggest that howls and barks serve separate functions. Howls appear optimized to convey information (i.e. data), while barks seem more suitable for attracting attention and acoustic ranging.

Characterizing noise in nonhuman vocalizations: Acoustic analysis and human perception of barks by coyotes and dogs

Measuring noise as a component of mammalian vocalizations is of interest because of its potential relevance to the communicative function. However, methods for characterizing and quantifying noise are less well established than methods applicable to harmonically structured aspects of signals. Using barks of coyotes and domestic dogs, we compared six acoustic measures and studied how they are related to human perception of noisiness. Measures of harmonic-to-noise-ratio (HNR), percent voicing, and shimmer were found to be the best predictors of perceptual rating by human listeners. Both acoustics and perception indicated that noisiness was similar across coyote and dog barks, but within each species there was significant variation among the individual vocalizers. The advantages and disadvantages of the various measures are discussed.

Testing Hypotheses of Bird Extinctions at Rio Palenque, Ecuador, with Informal Species Lists

Informally gathered species lists are a potential source of data for conservation biology, but most remain unused because of questions of reliability and statistical issues. We applied two alternative analytical methods (contingency tests and occupancy modeling) to a 35-year data set (1973-2007) to test hypotheses about local bird extinction. We compiled data from bird lists collected by expert amateurs and professional scientists in a 2-km(2) fragment of lowland tropical forest in coastal Ecuador. We tested the effects of the following on local extinction: trophic level, sociality, foraging specialization, light tolerance, geographical range area, and biogeographic source. First we assessed extinction on the basis of the number of years in which a species was not detected on the site and used contingency tests with each factor to compare the frequency of expected and observed extinction events among different species categories. Then we defined four multiyear periods that reflected different stages of deforestation and isolation of the study site and used occupancy modeling to test extinction hypotheses singly and in combination. Both types of analyses supported the biogeographic source hypothesis and the species-range hypothesis as causes of extinction; however, occupancy modeling indicated the model incorporating all factors except foraging specialization best fit the data.

Factors Influencing Occupancy Patterns of Eastern Newts across Vermont

Abstract Of the threats facing amphibian populations today, habitat transformation resulting from land use is among the most pressing. Although conservation of pond-breeding salamanders clearly requires protection of breeding ponds and their surrounding habitat, little is known about the effects of land use and other factors on the occurrence of salamanders in the dispersal/terrestrial phase of their life cycle. To determine these effects, we surveyed populations of Eastern Newts (Notophthalmus viridescens) at 551 stations across Vermont and modeled salamander distribution as a function of environmental variables hypothesized to influence site occupancy. We developed a set of 12 models based on seven a priori hypotheses of site occupancy. We hypothesized that occupancy was influenced by (1) amounts of available habitat types, (2) arrangement of these habitat types, (3) geographic position, (4) housing density, (5) road density, (6) short-term changes in habitat distribution, or (7) habitat structure at the stand level. We used a single-season occupancy model to rank and compare the 12 models. A total of 232 Eastern Newts was detected at 82 of 551 stations. Of the 12 models, amount of habitat within 0.5 km of the survey station best represented the field data. Strong effects were indicated for developed land (−), open water (+), and forest (+) cover. Given a survey station with average forest and open water characteristics, stations with >5% developed land classes within a 0.5-km buffer had a very low probability of occupancy. Further research is needed to determine the direct role of development on occupancy patterns.

Disentangling canid howls across multiple species and subspecies: Structure in a complex communication channel.

Wolves, coyotes, and other canids are members of a diverse genus of top predators of considerable conservation and management interest. Canid howls are long-range communication signals, used both for territorial defence and group cohesion. Previous studies have shown that howls can encode individual and group identity. However, no comprehensive study has investigated the nature of variation in canid howls across the wide range of species. We analysed a database of over 2000 howls recorded from 13 different canid species and subspecies. We applied a quantitative similarity measure to compare the modulation pattern in howls from different populations, and then applied an unsupervised clustering algorithm to group the howls into natural units of distinct howl types. We found that different species and subspecies showed markedly different use of howl types, indicating that howl modulation is not arbitrary, but can be used to distinguish one population from another. We give an example of the conservation importance of these findings by comparing the howls of the critically endangered red wolves to those of sympatric coyotes Canis latrans, with whom red wolves may hybridise, potentially compromising reintroduced red wolf populations. We believe that quantitative cross-species comparisons such as these can provide important understanding of the nature and use of communication in socially cooperative species, as well as support conservation and management of wolf populations.

Mob Mentality: Effect of a Mobbing Playback on Avian Detection Probabilities during Point Count Surveys

The probability of detecting an individual or species is an important parameter in studies using mark-recapture and occupancy models to estimate population sizes and occurrence. Because low detection probabilities result in biased estimators and decreased precision, biologists seek methods that maximize detection probability. We evaluated whether we could increase detections of bird species by playing a tape of Black-capped Chickadee (Poecile atricapillus) mobbing calls during point counts. We conducted trios of 10-minute counts (two pre-playback and a playback) at 684 stations throughout Vermont, in forested, agricultural/grassland, and developed habitats. For each of 73 species detected during the surveys, we used occupancy modeling and information-theoretic model selection and averaging methods to evaluate whether detection probabilities varied due to playback or habitat type. Models containing a playback effect accounted for over 90% of the Akaike weights for 41 species. With 15 of these species, habitat effects also accounted for over 90% of the Akaike weights. The playback increased estimated detection probability in all habitats for 14 species, decreased esti- mated detection probability for 20 species, and had an estimated effect that varied by habitat for 7 species (many species with habitat effects simply had differing magnitudes of the effect dependent on habitat). Smaller resident species were de- tected more often during tape playbacks, but responses were highly variable for most species and the responses did not appear to follow a taxonomic pattern. We encourage researchers to evaluate their list of target species carefully before de- ciding to use mobbing playbacks to enhance response rates; in many situations mobbing tapes will not enhance detections and may complicate the interpretation of model parameters.