Testing hypotheses in evolutionary ecology with imperfect detection: capture-recapture structural equation modeling

Studying evolutionary mechanisms in natural populations often requires testing multifactorial scenarios of causality involving direct and indirect relationships among individual and environmental variables. It is also essential to account for the imperfect detection of individuals to provide unbiased demographic parameter estimates. To cope with these issues, we developed a new approach combining structural equation models with capture-recapture models (CR-SEM) that allows the investigation of competing hypotheses about individual and environmental variability observed in demographic parameters. We employ Markov chain Monte Carlo sampling in a Bayesian framework to (1) estimate model parameters, (2) implement a model selection procedure to evaluate competing hypotheses about causal mechanisms, and (3) assess the fit of models to data using posterior predictive checks. We illustrate the value of our approach using two case studies on wild bird populations. We first show that CR-SEM can be useful to quantify the action of selection on a set of phenotypic traits with an analysis of selection gradients on morphological traits in Common Blackbirds (Turdus merula). In a second case study on Blue Tits (Cyanistes caeruleus), we illustrate the use of CR-SEM to study evolutionary trade-offs in the wild, while accounting for varying environmental conditions.     

Responses of pond-breeding amphibians to wildfire: short-term patterns in occupancy and colonization.

Wildland fires are expected to become more frequent and severe in many ecosystems, potentially posing a threat to many sensitive species. We evaluated the effects of a large, stand-replacement wildfire on three species of pond-breeding amphibians by estimating changes in occupancy of breeding sites during the three years before and after the fire burned 42 of 83 previously surveyed wetlands. Annual occupancy and colonization for each species was estimated using recently developed models that incorporate detection probabilities to provide unbiased parameter estimates. We did not find negative effects of the fire on the occupancy or colonization rates of the long-toed salamander (Ambystoma macrodactylum). Instead, its occupancy was higher across the study area after the fire, possibly in response to a large snowpack that may have facilitated colonization of unoccupied wetlands. Naive data (uncorrected for detection probability) for the Columbia spotted frog (Rana luteiventris) initially led to the conclusion of increased occupancy and colonization in wetlands that burned. After accounting for temporal and spatial variation in detection probabilities, however, it was evident that these parameters were relatively stable in both areas before and after the fire. We found a similar discrepancy between naive and estimated occupancy of A. macrodactylum that resulted from different detection probabilities in burned and control wetlands. The boreal toad (Bufo boreas) was not found breeding in the area prior to the fire but colonized several wetlands the year after they burned. Occupancy by B. boreas then declined during years 2 and 3 following the fire. Our study suggests that the amphibian populations we studied are resistant to wildfire and that B. boreas may experience short-term benefits from wildfire. Our data also illustrate how naive presence-non-detection data can provide misleading results.     

Landscape history, fragmentation, and patch occupancy: models for a forest bird with limited dispersal.

We developed and tested patch occupancy models for an endemic understory bird with limited dispersal ability, the Chucao Tapaculo (Scelorchilus rubecula), in two South American temperate rain forest landscapes that differed in levels and duration of forest loss. We assessed cover changes since 1961 in each landscape and surveyed patches for Chucao Tapaculo occupancy. We then developed incidence-based predictive models independently for each landscape and tested each model reciprocally in the alternative study area. We thereby assessed the domain of model applicability and identified those predictor variables with general effects and those that varied between the two landscapes. The two models were consistent regarding variable selection, and predictive accuracy of each model was high in the landscape where training data were collected. However, the models differed substantially in the magnitudes of effects related to patch size, with larger unoccupied patches observed in the landscape with the more advanced stage of fragmentation. Due to this discrepancy, each model performed poorly when applied to the alternative landscape, potentially reflecting the contrasting stages of habitat loss. Although it was impossible to dissociate effects of level and duration of forest loss, we viewed the landscapes as representing two extremes along a continuum of fragmentation, providing insights into potential trajectories for portions of the biome where deforestation is occurring. Further, our data suggest that static equilibrium models developed from distribution patterns in recently fragmented landscapes may overestimate persistence when used as a forecasting tool, or when extrapolated to alternative landscapes where fragmentation is more advanced. Thus, we suggest that landscapes used as standards for model building should be selected with caution. We recommend that distribution patterns be obtained from landscapes where fragmentation is advanced, preferably with histories of fragmentation long enough that time-delayed extinctions already would have occurred.     

Male mate guarding in a socially monogamous mammal,the round-eared sengi: on costs and trade-offs

Mate guarding is predicted to be one of the driving forces for the evolution of monogamy, but supporting evidence in free-living mammals is rare. The goals of our study were three-pronged. First, we tested if mate guarding, measured as intrapair distance, occurs as a behavioral tactic in round-eared sengis (Macroscelides proboscideus), a socially monogamous species lacking paternal care and in which females breed asynchronously, producing 2–3 litters during an 8-month long breeding season. Second, we determined if mate guarding involves costs which we identified as changes in male body mass. Third, we investigated whether variation in individual investment in mate guarding depended on the males’ body mass and the number of neighboring males. Field data were collected in a semidesert in South Africa using radio-tracking, trapping, and direct observations during three successive breeding seasons. Mate guarding strongly depended on the females’ reproductive state, and all males started to guard their mates prior to and during estrus, as exemplified by reduced intrapair distance. Mate guarding incurred costs: overall, males lost about 5% of body mass. Male body mass loss and initial body mass were negatively related to the intensity of precopulatory mate guarding. Furthermore, during estrus intrapair distance was inversely correlated with the number of neighboring males. The results show that mate guarding is the predominant male tactic in round-eared sengis. However, since mate guarding imposed costs, males may balance benefits and costs associated with guarding by varying their effort in relation to their physical capabilities and the competitive environment.     

Fruit detection and discrimination by small fruit-eating bats (Phyllostomidae): echolocation call design and olfaction

We studied the role of echolocation and other sensory cues in two small frugivorous New World leaf-nosed bats (Phyllostomidae: Artibeus watsoni and Vampyressa pusilla) feeding on different types of fig fruit. To test which cues the bats need to find these fruit, we conducted behavioral experiments in a flight cage with ripe and similar-sized figs where we selectively excluded vision, olfaction, and echolocation cues from the bats. In another series of experiments, we tested the discrimination abilities of the bats and presented sets of fruits that differed in ripeness (ripe, unripe), size (small, large), and quality (intact(infested with caterpillars). We monitored the bats' foraging and echolocation behavior simultaneously. In flight, both bat species continuously emitted short (<2 ms), multi-harmonic, and steep frequency-modulated (FM) calls of high frequencies, large bandwidth, and very low amplitude. Foraging behavior of bats was composed of two distinct stages: search or orienting flight followed by approach behavior consisting of exploration flights, multiple approaches of a selected fruit, and final acquisition of ripe figs in flight or in a brief landing. Both bat species continuously emitted echolocation calls. Structure and pattern of signals changed predictably when the bats switched from search or orienting calls to approach calls. We did not record a terminal phase before final acquisition of a fruit, as it is typical for aerial insectivorous bats prior to capture. Both bat species selected ripe over unripe fruit and non-infested over infested fruit. Artibeus watsoni preferred larger over smaller fruit. We conclude from our experiments, that the bats used a combination of odor-guided detection together with echolocation for localization in order to find ripe fruit and to discriminate among them.     

Necromass production: studies in undisturbed and logged Amazon forests.

Necromass stocks account for up to 20% of carbon stored in tropical forests and have been estimated to be 14-19% of the annual aboveground carbon flux. Both stocks and fluxes of necromass are infrequently measured. In this study, we directly measured the production of fallen coarse necromass (> or = 2 cm diameter) during 4.5 years using repeated surveys in undisturbed forest areas and in forests subjected to reduced-impact logging at the Tapajos National Forest, Belterra, Brazil (3.08 degrees S, 54.94 degrees W). We also measured fallen coarse necromass and standing dead stocks at two times during our study. The mean (SE) annual flux into the fallen coarse necromass pool in undisturbed forest of 6.7 (0.8) Mg x ha(-1) x yr(-1) was not significantly different from the flux under a reduced-impact logging of 8.5 (1.3) Mg x ha(-1) x yr(-1) With the assumption of steady state, the instantaneous decomposition constants for fallen necromass in undisturbed forests were 0.12 yr(-1) for large, 0.33 yr(-1) for medium, and 0.47 yr(-1) for small size classes. The mass weighted decomposition constant was 0.15 yr(-1) for all fallen coarse necromass. Standing dead wood had a residence time of 4.2 years, and approximately 0.9 Mg x ha(-1) x yr(-1) of this pool was respired annually to the atmosphere through decomposition. Coarse necromass decomposition at our study site accounted for 12% of total carbon remineralization, and total aboveground coarse necromass was 14% of the aboveground biomass. Use of mortality rates to calculate production of coarse necromass leads to an underestimation of coarse necromass production by 45%, suggesting that nonlethal disturbance such as branch fall contributes significantly to this flux. Coarse necromass production is an important component of the tropical forest carbon cycle that has been neglected in most previous studies or erroneously estimated.     

Colonization and extinction in dynamic habitats: an occupancy approach for a Great Plains stream fish assemblage

Despite the importance of habitat in determining species distribution and persistence, habitat dynamics are rarely modeled in studies of metapopulations. We used an integrated habitat-occupancy model to simultaneously quantify habitat change, site fidelity, and local colonization and extinction rates for larvae of a suite of Great Plains stream fishes in the Arikaree River, eastern Colorado, USA, across three years. Sites were located along a gradient of flow intermittency and groundwater connectivity. Hydrology varied across years: the first and third being relatively wet and the second dry. Despite hydrologic variation, our results indicated that site suitability was random from one year to the next. Occupancy probabilities were also independent of previous habitat and occupancy state for most species, indicating little site fidelity. Climate and groundwater connectivity were important drivers of local extinction and colonization, but the importance of groundwater differed between periods. Across species, site extinction probabilities were highest during the transition from wet to dry conditions (range: 0.52-0.98), and the effect of groundwater was apparent with higher extinction probabilities for sites not fed by groundwater. Colonization probabilities during this period were relatively low for both previously dry sites (range: 0.02-0.38) and previously wet sites (range: 0.02-0.43). In contrast, no sites dried or remained dry during the transition from dry to wet conditions, yielding lower but still substantial extinction probabilities (range: 0.16-0.63) and higher colonization probabilities (range: 0.06-0.86), with little difference among sites with and without groundwater. This approach of jointly modeling both habitat change and species occupancy will likely be useful to incorporate effects of dynamic habitat on metapopulation processes and to better inform appropriate conservation actions.     

Comparison of a two-stage sampling design and its composite sample alternative: An application to soil studies

The objective of a long-term soil survey is to determine the mean concentrations of several chemical parameters for the pre-defined soil layers and to compare them with the corresponding values in the past. A two-stage random sampling procedure is used to achieve this goal. In the first step, n subplots are selected from N subplots by simple random sampling without replacement; in the second step, m sampling sites are chosen within each of the n selected subplots. Thus n · m soil samples are collected for each soil layer. The idea of the composite sample design comes from the challenge of reducing very expensive laboratory analyses: m laboratory samples from one subplot and one soil layer are physically mixed to form a composite sample. From each of the n selected subplots, one composite sample per soil layer is analyzed in the laboratory, thus n per soil layer in total. In this paper we show that the cost is reduced by the factor m — 1 when instead of the two-stage sampling its composite sample alternative is used; however, the variance of the composite sample mean is increased. In the case of positive intraclass correlation the increase is less than 12.5%; in the case of negative intraclass correlation the increase depends on the properties of the variable as well. For the univariate case we derive the optimal number of subplots and sampling sites. A case study is discussed at the end.     

Detecting Rare Species with Random or Subjective Sampling: a Case Study of Red-Listed Saproxylic Beetles in Boreal Sweden

Abstract:  Efficient sampling design in field studies is important for economical and statistical reasons. We compared two ways to distribute sampling effort over an area, either randomly or subjectively. We searched for red-listed saproxylic (wood-living) beetles in 30 spruce stands in boreal Sweden by sifting wood from dead trees. We randomly selected positions within each stand with a geographic positioning system and sampled the nearest dead tree (random sample). In the same stand we also sampled dead trees that, based on literature, were likely to host such species (subjective sampling). The subjective sampling (two to five samples per stand, depending on stand size) was compared with the higher, random sampling effort (fixed level of 12 samples/stand). Subjective sampling was significantly more efficient. Red-listed species were found in 36% of the subjective samples and in 16% of the random samples. Nevertheless, the larger random effort resulted in a comparable number of red-listed species per stand and in 13 detected species in total (vs. 12 species with subjective sampling). Random sampling was less efficient, but provided an unbiased alternative more suitable for statistical purposes, as needed in, for example, monitoring programs. Moreover, new species-specific knowledge can be gained through random searches.     

Predator detection and avoidance: Ornamentation of tube-caps of Diopatra spp. (Polychaeta: Onuphidae)

One strategy for predator avoidance involves the creation of a structural refuge. Onuphid polychaetes characteristically ornament above-sediment portions of their tubes (=tube-caps) with shell and algal debris. These species feed on the sediment surface through an opening in the tube-cap, and thus the ability to detect a surface predator while feeding would be advantageous. An investigation of the function of the ornamentation in Diopatra spp. suggests that ornamentation facilitates predator detection and avoidance. Three intensities of mechanical disturbance were applied directly to D. ornata tube-caps. When ornamented tube-caps were stimulated, the response of worms to the three intensities were significantly different, and increased in duration with greater intensities. In contrast, when no ornamentation was present, the responses were not significantly different, and were similar to the low-intensity response when ornamentation was present. This suggests that ornamentation should allow a worm to distinguish between harmful (high intensity due to mobile epifaunal predators) and profitable (low intensity due to drift algae) disturbances, and furthermore, worms with ornamented tube-caps should be more successful in escaping surface predators. Densities of intertidal populations of D. cuprea at Tom's Cove, Virginia, USA, correlated with the amount of tube ornamentation, consistent with this predator detection and avoidance hypothesis. Final tube-cap lengths of laboratory D. ornata and field D. cuprea were inversely related to the size of attached debris. When large debris was attached, cap formation ceased earlier and caps were shorter than when small debris or no debris was attached. Cryptic and food-catching functions would predict that highly ornamented tubes would be most advantageous, while only a few large debris would be required for disturbance transmission. Laboratory specimens showed no selectivity between 0.5 or 1.5 cm² shell; or 1.0 cm² or 3 to 8 cm² algae, and utilized shell and algae when available and according to relative abundance. Tube-caps of field specimens also showed positive correlations between shell attached and shell abundance in the local sediment. While such lack of selectivity may enhance cryptic properties of the tube-cap, it is argued that conditions seldom exist which would permit selectivity of debris size or of specific material. These data are consistent with the hypothesis that ornamentation functions as a created refuge for predator detection and avoidance and further suggest that the availability of ornamental debris in the environment indirectly can affect these species' distributions and abundances.     

Incorporating species relocation in reserve design models: An example from Ft. Benning GA

Given the increasing expansion of human dominated landscapes it often becomes necessary to relocate endangered and at-risk species from existing habitat areas. Further, there is growing research stating that climatic and atmospheric changes attributed to climate change are already affecting species distributions and geographic ranges, requiring endangered species to be relocated. Existing reserve design models mostly focus on choosing the optimal land area given existing species distributions and have not incorporated species relocation as a criterion. This paper introduces linear integer programming formulations for the relocation of multiple populations of a species at risk to clustered conservation areas. We present a basic clustered relocation model and extend the model to minimize the distances of relocation. We apply the models to a dataset related to Gopher Tortoise (GT), a keystone species currently considered ‘at risk’, at Ft. Benning Georgia where expanding military training needs require the relocation of GTs. We present the results and discuss the trade-off between compactness and relocation considerations using an efficiency frontier.     

Metachronal swimming in Antarctic krill: gait kinematics and system design

Metachronal swimming, in which adjacent appendages stroke in sequence, is widespread among crustaceans inhabiting the transitional flow realm in which both viscosity and inertia effects are important. However, the design and operation of this propulsion system in response to various hydrodynamic, energetic, and behavioral needs have not been well investigated. We examine free-swimming Antarctic krill (Euphausia superba) as a model species and identify three distinct behavioral swimming gaits. The pleopod kinematics of these gaits, hovering, fast-forward swimming, and upside-down swimming, are quantified via image analysis of high-speed video. Pleopod stroke amplitude and frequency were found to vary significantly among these swimming modes. In order to increase swimming speed, krill were found first to increase stroke amplitude and secondarily to increase beat frequency. The kinematics of these distinct swimming modes provide insight as we consider multi-appendage metachronal swimming from a design standpoint. The ratio of the distance between adjacent appendage bases and appendage length is identified as a key parameter in metachrony, the value of which is constrained to a narrow range for a wide variety of species.     

Evolution of Cardium (Cerastoderma) edule,C. lamarcki and C. glaucum: studies of DNA-variation

Genetic relations between populations of Cardium (Cerastoderma) edule (L.) and C. lamarcki (Reeve) from Danish waters and C. glaucum (Bruguiére) from the Mediterranean Sea were studied using DNA analysis. Repetitive sequences of nuclear DNA were compared by Southern blotting of enzyme digested DNA fragments, separated on agarose gel by electrophoresis. Fragments were visualized by autoradiography using ³²P labelled probes. These were nick translated, randomly sheared fragments of the whole chromosomal genomes of two species of distant taxa — Mytilus edulis and Mya arenaria, respectively. The number of detected DNA bands was 18 for C. edule, 32 for C. lamarcki, and 24 for C. glaucum. Measurements of genetic distances, based on comparisons of these repetitive sequences using a simple dissimilarity coefficients, show the greatest distance, i.e., 0.84 between C. edule and C. lamarcki, 0.69 between C. edule and C. glaucum, and the smallest, 0.52 between C. lamarcki and C. glaucum. Comparisons of the three Cardium types, using gene spliced and cloned C. edule nuclear DNA (nDNA) fragments as probes, accorded with these findings. Also, analysis of mitochondrial DNA (mtDNA) from eggs and hepatopancreas, although not in itself conclusive, accorded with the results of the analyses of nDNA. MtDNA was measured by electron micrograph to 15 000 base pairs (bp), and sizes of restriction enzyme digested fragments were compared by gel electrophoresis. The results of the comparisons based on repetitive sequences supported by the two other studies suggest that C. edule resembles the ancestral form more than the two others, and that C. lamarcki is the most different and recently evolved form. This hypothesis is discussed in relation to the geographical distribution of the species complex.     

Market design in cap and trade programs: Permit validity and compliance timing

Cap and trade programs have considerable heterogeneity in permit validity and compliance timing. For example, permits have different validities across time (e.g., banking, borrowing, and seasons) and space (e.g., zonal restrictions), and compliance timing can be annual, in overlapping cycles, or in multi-year periods. We discuss nine prominent cap and trade programs along these dimensions and construct a general model of permit validity and compliance timing. We derive sufficient conditions under which abatement is invariant to compliance timing, i.e., compliance timing cannot smooth abatement cost shocks. Under these conditions, (i) expected compliance costs are invariant, (ii) the variance of compliance costs increases with the delayed compliance, (iii) equilibrium prices may not be unique, and (iv) the delayed compliance equilibrium may rely on “degenerate” prices not determined by marginal abatement costs. We demonstrate the model's broad applicability by illustrating different types of temporal permit validity.     

Biodiversity conservation through faith and tradition in India: some case studies

SUMMARY

In India, since ancient times, several members of the flora and fauna have been identified with particular personalities of the Hindu pantheon and are worshipped. Many wild animals, such as tigers, lions and elephants, and birds such as the peacock, owl and swan have been revered as the vehicles of Gods and Goddesses and worshipped. Several trees, such as mango, coconut, khejri, emblica, tamarind, ashoka, madhuca and wood apple, and herbs such as basil are considered highly sacred and worshipped. This belief and these practices continue in one form or another even today in certain sections of Indian society, particularly the rural folk and the aboriginal tribes living in the forest. Nature worship is a form of tribal belief and the faith of these laymen has helped to conserve many natural ecosystems in India. They have preserved many pristine forests—called ‘sacred groves’ in their original forms. Sacred groves are rich in biodiversity. They are the last refuge of the rare flora and fauna which have otherwise disappeared from the modern world. Sacred groves have become part of the ‘biosphere reserves’ of India.     

Space-time Bayesian small area disease risk models: development and evaluation with a focus on cluster detection

This paper extends the spatial local-likelihood model and the spatial mixture model to the space-time (ST) domain. For comparison, a standard random effect space-time (SREST) model is examined to allow evaluation of each model’s ability in relation to cluster detection. To pursue this evaluation, we use the ST counterparts of spatial cluster detection diagnostics. The proposed criteria are based on posterior estimates (e.g., misclassification rate) and some are based on post-hoc analysis of posterior samples (e.g., exceedance probability). In addition, we examine more conventional model fit criteria including mean square error (MSE). We illustrate the methodology with a real ST dataset, Georgia throat cancer mortality data for the years 1994–2005, and a simulated dataset where different levels and shapes of clusters are embedded. Overall, it is found that conventional SREST models fair well in ST cluster detection and in goodness-of-fit, while for extreme risk detection the local likelihood ST model does best.     

Role of competition and predation in determining habitat occupancy of Cerithiidae (Gastropoda: Prosobranchia) on the rocky,intertidal, Red Sea coasts of Sinai

Six cerithiid snail species occur on rocky intertidal flats along the Sinai coasts of the Red Sea: Clypeomorus moniliferum, Cerithium caeruleum, C. scabridum, C. columna, Clypeomorus tuberculatum and Cerithium echinatum. The present study, conducted at 22 stations covering almost the whole length of Sinai, covered the 4 yr period from October 1972 to August 1976, and describes the habitat occupancy of 5 of these species (C. echinatum is excluded for lack of data). Several of these species (sometimes all) often occur together, and in such cases are distincly segregated by habitat. However all species considerably overlap in their distribution along the axes of four major interrelated abiotic gradients, thus excluding the possibility that habitat segregation is determined by larval settlement preferences. Other distributional patterns observed at some sites, such as lack of overlap or contact between belts of the various species and the relative abundance of food available to all species, make postlarval competitive interactions unlikely. The existence and the degree of proximity of a coral reef with its associated predatory fishes, influence the cerithiids' distributional patterns. Differences between the cerithiid species in their vulnerability to fish predation, associated with differences between sites in the abundance and the accessibility of predatory fishes, and in the availability of refuges for each cerithiid species, can satisfactorily explain the observed distributional patterns including co-occurrence with habitat segregation. It is proposed that habitat segregation is caused by predation on young stages by generalist fishes which may totally eliminate a certain species at a given site; the same site may provide refuges for recruits of another species, allowing these to survive to an advanced age. In general, the flat's structural complexity is associated with its diversity of refuges from predation, and hence with the number of co-existing species. This mechanism for co-existence and habitat segregation in tropical Cerithiidae may also be instrumental in maintaining the high species diversity of other tropical benthic communities.Paper No. 12 in the series Colonization of the Eastern Mediterrancan by Red Sea species immigrating through the Suez Canal