William B. Krohn

Predicting occurrence of river otters by habitat on Mount Desert Island, Maine.

We developed a model predicting occurrence of river otters (Lutra canadensis) by comparing 19 watersheed used by otters with 14 unused watersheds during 1985-87. Of 39 habitat variables initially measured, 4 provided good discrimination between used and unused watershed. Otter use was negatively associated with the proportion of mixed hardwood-softwood stands in forested areas adjacent to waterways and positively associated with the number of beaver flowages, watershed length, and average shoreline diversity. Our model correctly classified 94% of the watersheds and demonstrated 88% better classification than could be achieved by chance alone

MODELING THE OCCURRENCE OF BIRD SPECIES: ARE THE ERRORS PREDICTABLE?

Typical assessments of models where many species occurrences are predicted (e.g., from species–habitat matrices or Gap Analyses) report overall omission and commission errors. Yet species’ attributes suggest that we may predict a priori that some species are more likely to be modeled correctly than others. Because the likelihood of modeling species correctly is related to species incidences in surveys, a method was created that ranked the 183 avian species known to be breeding in Maine as to how likely they would be to occur in surveys. Attributes (e.g., population level, niche width, aggregation) were used to model 79% of the variation in incidence within the Maine Breeding Bird Atlas. Likelihood of Occurrence Ranks (LOORs) were assigned to each species based upon the modeled incidences to reflect how likely the species are to be observed in future surveys. The occurrence of birds on areas with species checklists were then modeled and compared to the LOORs. For five of six areas, the number of species correctly modeled using species–habitat associations was highly correlated with LOORs: species judged a priori to be likely to be modeled correctly actually were. For one large area (9172 ha) with a checklist covering 52 years, the number of species correctly modeled was not correlated with LOORs, evidence that the checklist is essentially complete. In general, sites with checklists from many years (e.g., >10 yr) and from large areas (e.g., >1000 ha) yielded the lowest commission error. These results demonstrate that the confidence assigned to results where the occurrences of species are modeled (e.g., Gap Analysis) is highly dependent on the test sets and the species modeled.

Dispersal of juvenile fishers in Maine

We studied natal dispersal of fishers (Martes pennanti) in a harvested population in southcentral Maine during 1984-90 because of concern over the high level of harvest and a lack of information about fisher dispersal. Probability that an individual would disperse by the end of its first year was 73 and 100% for males and females, respectively (n = 21 M, 12 F). However, 32% of males and no females dispersed before the age of 9 months. Mean distance between natal and adult home ranges did not differ (t = 0.16, P = 0.88) between males (n = 8, x = 10.8 km, range = 4.1-19.5) and females (n = 5, x = 11.2 km, range = 5.0-18.9). Dispersal, in combination with intrasexual territoriality of adults, may limit fisher density and competition for food

HABITAT MODELS FOR NESTING BALD EAGLES IN MAINE

We derived habitat models for nesting bald eagles (Haliaeetus leucocephalus) in river, lake, marine mainland, and marine island habitats. We measured 39 variables and used discriminant analysis to contrast 82 occupied nest sites and 88 random sites in Maine. Compared to random sites, bald eagle nests were located on river stretches with a larger basin area, less forest edge, and closer to the shore. The lake model indicated that bald eagle nests were positively associated with the number of superdominant trees and negatively associated with distance to water, area of land disturbed by humans, and area of land harvested for timber. The number of diadromous fish species and area of water <1.8 m deep at low tide were positively associated with bald eagle nests in marine mainland habitats; length of roads near nests was a negative variable. Nests were located on smaller marine islands than were random points (P < 0.005), resulting in nesting islands characterized by larger terrestrial and aquatic openings in the forest canopy, less forest edge, and a smaller area of combined shallow water and intertidal area. Correct classification rates for models varied from 75% for the marine island model to 100% for the river model. Similar classification rates, which never differed by more than 8%, occurred during cross validation. Validity of marine models was further suggested by the establishment of 4 bald eagle nests 51.9 km from random sites classified as nest sites; no nests have been established ?1.9 km from random sites classified as unused.

Predicting broad-scale occurrences of vertebrates in patchy landscapes

Spatially explicit landscape-scale models that predict species distributions, where patches of habitat are shown as having potential to be occupied or unoccupied, are increasingly common. To successfully use such data, one should understand how these predicted distributions are created and how their relative accuracies are assessed. Geographic ranges, defined upon observations (e.g., atlases), literature review, and expert review, are a primary data layer. A map of land cover is created, often from interpretation of satellite imagery or other remotely-sensed data. Species/habitat associations are defined based upon a literature review and expert review, describing associations for habitats derived from the cover map. Included as ancillary associations are how species relate to physical features, where appropriate, such as elevation and hydrography. The three layers of information (range, land cover, and associations) are merged, often using raster-based algebraic statements that exclude unused habitats or patches outside the range of a species. The accuracy of predictions for a suite of species is typically assessed with surveys by comparing the species predicted to occur in an area to the species observed. Omission (i.e., present in species lists but not predicted) and commission (i.e., predicted but not present in lists) errors are reported. Errors may be due to many sources. For example, ranges of species change, cover types may be misidentified, species/habitat associations may be incorrect or change, or species may be rare and unlikely to be seen in surveys and judged in-error even though the species may be present. An example is given of an appropriate use of broad-scale species predicted distributions, in which patterns and threats to Maine terrestrial vertebrate diversity are summarized.

Activity patterns, movements, and reproductive ecology of fishers in southcentral Maine

Most activity of adult fishers ( Martes pennanti ) occurred shortly before sunrise and after sunset, but some activity occurred throughout the day and night. Males and females showed similar amounts of activity, and both were active more frequently in summer than winter. Movements of males were greatest during the spring breeding season; nonreproductive females moved similar distances during all seasons. Eleven maternal dens were in tree cavities; the denning period lasted 8–12 weeks from March through May. The maximum annual parturition rate among females ≥2 years old was 75% ( n = 4). This was less than pregnancy rates indicated by corpora lutea, and suggested that some pregnancies were unsuccessful. We hypothesize that delayed implantation allows mating to occur at a favorable time for adults, and births to occur when offspring have the maximum time for development before winter.

QUANTITATIVE DELINEATION AND CHARACTERIZATION OF HIERARCHICAL BIOPHYSICAL REGIONS OF MAINE

Natural resource managers and researchers are often faced with the issue of how to subdivide large landscapes into smaller, homogeneous units for analytical or administrative purposes. We used data in a Geographic Infor- mation System and cluster analyses to divide Maine into broad-scale regions based on elevation, slope, two measures of species richness, and three weather variables. The locations and numbers of these biophysical regions were affected by variable selection, cell size (and sampling), and euclidean distances used to terminate clustering. We present rationales, largely based on our intended uses of these regions, for selecting the seven variables, a cell size (21.44 km2), and a euclidian cut-off (1.56). Thirteen biophysical regions, in fourteen units, were delineated and characterized; the regions were hierarchically arranged in five levels. Five of the seven variables used in the final clustering were abiotic versus biotic, suggesting that regions relate to potential as well as current ecological conditions. The euclidian distances defining individual regions can be used to quantify the biophysical distinctness of each region (and grid-cell).

RECENT AND HISTORICAL DISTRIBUTIONS OF CANADA LYNX IN MAINE AND THE NORTHEAST

Abstract The contiguous United States population of Canada lynx (Lynx canadensis Kerr) is listed as threatened under the federal Endangered Species Act. However, the historic distribution of lynx in the Northeast is poorly understood. We used museum records, bibliographic records, and interviews to reconstruct the past distribution of lynx in Maine, which is at the current southern limit of the species distribution in the eastern United States. We found a total of 118 records, representing at least 509 lynx in Maine. Lynx were observed throughout Maine, 1833–1912, with the exception of coastal areas. After 1913, lynx were most common in the forests of western and northern Maine, and absent to rare along the coast, but had not returned to southern Maine by 1999. Thirty-nine kittens representing at least 21 litters were distributed throughout northern and western Maine, 1864–1999. Populations apparently fluctuated, and in some years 200–300 lynx were harvested in Maine. Prior to the 1900s, lynx were much more widely distributed in the Northeast, ranging from Pennsylvania north into Quebec. Because Canada lynx have had a long presence in northern New England, and at times were relatively common, this species merits serious consideration in conservation planning in this region.

Winter use of glyphosate-treated clearcuts by moose in Maine

Aerial treatment of naturally regenerating clearcuts with the herbicide glyphosate initially reduces the availability of deciduous browse, but may subsequently improve bedding cover for moose (Alces alces). However, the potential effects of these vegetative changes on use of clearcuts by moose has received little study. We studied effects of glyphosate treatment of clearcuts in Maine on (1) use of clearcuts by moose and (2) conifer cover during 2 periods, 1-2 and 7-11 years posttreatment. We made counts of moose tracks, beds, and pellet groups on transects in treated and untreated clearcuts in January-March 1992 and 1993 and measured conifer densities in January-March 1991-93. At 1 and 2 years posttreatment, tracks of foraging moose were 57 and 75% less abundant on treated than untreated clearcuts (P = 0.013). Counts of moose beds, total tracks, and pellet groups exhibited similar patterns as tracks of foraging moose but did not differ (P > 0.1) between treatments. At 7-11 years posttreatment, tracks of foraging moose (P = 0.05) and moose beds (P = 0.06) were greater on treated than untreated clearcuts. Conifer densities at 1-2 years posttreatment were not affected (P > 0.1) by treatment, but conifers 2.0-2.9 m tall were 2 times more abundant (P < 0.1) on treated than untreated clearcuts at 7-11 years posttreatment. Less foraging activity at 1-2 years posttreatment appeared to be the result of reduced browse availability because conifer cover for bedding was similar on treated and untreated clearcuts. We hypothesized that greater counts of tracks of foraging moose on older treated clearcuts was due to increased foraging activity on sites with more abundant conifer cover.

Martes Foot-Loading and Snowfall Patterns in Eastern North America

American martens (Martes americana) and fishers (M. pennanti) require large areas and live in complex, interacting communities of medium and large size carnivores. Nevertheless, habitat studies of these species continue to emphasize mid-to fine-scale habitat relationships, and rarely examine interspecific relations. Based on our data on foot-loading (ratio of body mass to total foot area) for fishers and martens, and snowfall patterns across eastern north america, we conclude that broader-scale habitat and interspecific relations of these 2 species may affect their regional distributions. Although foot-loading was influenced by sex (P 2 times greater (P<0.001) than martens across all of the 4 age-sex classes. Relative to other large- and medium-sized carnivores in the forests of eastern North America, the 2 Martes species have the shortest legs, and thus are most dependent on low foot-loading for mobility in soft snow. To assess temporal and spatial variation in snowfall as related to potential Martes distributions, we used a snowfall threshold reported for maine to define the mid-point of a zone with overlapping populations of fishers and martens and applied this threshold (240 cm mean annual snow) to regional snowfall data for 1970–90. Regression analyses of weather data in conjunction with data on latitude, longitude, and elevation were used to model mean annual snowfall. Since the late 1700’s, there has been a general warming trend across eastern North America. If previously proposed hypotheses that snow limits fishers, and large populations of fishers limit martens are true, then one would predict that martens historically occurred south of where they do today. Further, if snowfall continues to decline in the region, fisher populations may expand and martens may decline. To test these predicted broad-scale distribution patterns, we suggest that past and modern occurrence data for fishers, martens, and other forest carnivores be examined across the historic range of both species to evaluate the hypothesis that interactions among morphology and climate affect distribution and degree of sympatry in North American Martes.