Richard M. DeGraaf

Exploring the ecology of suburban wildlife

The fringes of cities, and rural properties within commuting distance of cities, experience some of the highest rates of development in the world. This can cause dramatic changes to the landscape, the alteration of ecological functions, and a reduction in biodiversity. With the spread of suburbia, however, come opportunities for some species to exploit new resources. While many wild creatures can enrich the lives of suburban dwellers, large increases in the populations of species such as deer, beaver, and coyotes can lead to a change in status from resource to pest. For several decades, wildlife managers have alternately embraced and ignored issues related to urban and suburban wildlife. Today, management of suburban areas challenges wildlife agencies on two fronts: the threat to habitat and biodiversity and the problem of “overabundant” wildlife. This is not only a tremendous management challenge, but also an educational opportunity to help people understand the natural world and their place in it.

Bird species diversity and nesting success in mature, clearcut and shelterwood forest in northern New Hampshire, USA

Bird species distribution and predation rates on natural and artificial nests were compared among unmanaged mature, shelterwood, and clearcut northern hardwoods forest to evaluate the effect of these practices on bird populations. Twenty-three of the 48 bird species detected during the study differed significantly in abundance among unmanaged mature forest, shelterwoods, and clearcuts. Results of multiple regressions of bird abundance and habitat variables suggest that differences in bird species distribution among treatments were the result of differences in habitat structure among treatments. Bird species diversity and species richness were significantly higher in shelterwoods than either mature forest or clearcuts, although there were bird species that occurred exclusively, or nearly so, in each of the three treatments. Predation rates on artificial nests were lowest in mature forest, and predation rates on natural nests was highest in mature forest, although neither of these differences was statistically significant. We conclude that use of partial cutting exclusively would result in the decline of several species of mature forest and clearcut specialists, and, consequently, a decrease in species diversity at the landscape scale. The use of a variety of silvicultural techniques is recommended to maintain bird species diversity in forested landscapes.

Avian guild structure and habitat associations in suburban bird communities

Breeding bird communities were compared in three suburbs: a 70-year-old area of large houses along streets shaded by mature trees (MT), primarily oaks (Quercus) and elms (Ulmus); a 15-year-old area built upon open agricultural land with young maple (Acer) street trees (YT); and a 15-year-old area on which houses were built in small clearings within a second-growth oak-pine (Quercus-Pinus) woodland (OP). Bird censuses each year for 5 years revealed that YT supported the lowest total avian density of the three suburbs; OP supported the greatest variety and total density of insectivores, and the lowest number of ground-gleaning omnivores; and MT supported the highest total avian density, comprised primarily of ground-foraging seed eaters and omnivores. Among nesting guilds, OP contained the fewest ground/herb nesters, and MT the most. Also, OP had the fewest shrub nesters, and MT the most. Tree cavity and twig nesters were significantly more numerous in OP, and tree branch nesters fewer in YT than in either OP or MT. Analyses of habitat structure revealed that shrub maturity is more important than numbers of shrubs, and that planted trees, no matter how mature or abundant, do not replace natural forest stands as habitat for most insectivorous species.

Predation on artificial nests in forested riparian buffer strips

We used artificial nests to examine predation in riparian buffer strips created by commercial clear-cutting and in unharvested control areas on industrial forestlands in eastern Maine. Nests in riparian buffer strips were depredated more often than those in intact riparian forests. This pattern was similar for both ground and shrub nests and for both trials. Predation rate for nests in control stands was 15%, compared to 31% in 20-40m wide buffer strips along tributary streams (P = 0.016) and 23% in 60-80m wide buffer strips along mainstem streams (P = 0.045). Predation rates were similar (P = 0.41) in mainstem and tributary buffer strips. Greater predation rates documented for nests in riparian buffer strips likely resulted from an elevated number and diversity of predators associated with the narrow, linear forest stands. Remotely-triggered cameras placed on a subset of nests revealed 6 species of nest predators. Predators identified at nests were mostly forest species and not species directly associated with riparian habitats. Red squirrels (Tamiasciurus hudsonicus) and blue jays (Cyanocitta cristata) were responsible for >50% of the identified depredations. Black bears (Ursus americanus) were photographed only in tributary buffer strips and may have been using them to travel between larger forested stands. Increased predation of eggs and young probably reduces the nesting success of birds in riparian buffer strips. Managers should leave wide (≥ 150-m) buffer strips along riparian zones to reduce edge-related nest predation, especially in landscapes where buffer strips are an important component of the existing mature forest.

Associations between breeding bird abundance and stand structure in the White Mountains, New Hampshire and Maine, USA

Assessment of faunal distribution in relation to landscape features is becoming increasingly popular. Technological advances in remote sensing have encouraged regional analyses of the distributions of terrestrial vertebrates. Comparisons of the strength of association of habitat characteristics at various scales of measurement of habitat structure are rare. We compared the associations of forest cover-type, stand size-class, and stand structure to abundance of breeding bird species in managed forest in northern New England. We surveyed breeding birds and measured stand structure in 20 stands to test the hypothesis that forest cover-type, stand size-class, and structure variables were equally associated with numbers of forest birds. We fit regression models to data from each data source to predict the log number of individuals for each species. We restricted our analyses to cover-types with > 1 size-class and to size-classes representing > 1 cover-type, and restricted our comparisons to bird species with at least 10 observations/yr for 2 yr. Of 31 bird species that met our criteria for analysis, a significant (P < 0.05) association was detected between bird abundance and structure data for 30 species, cover-type data for 19 species, and size-class data for 10 species. Stand structure was the best predictor of bird abundance for 25 species, cover-type for 5 species, and size-class for none. Of the 14 structure variables used in the analyses, total foliage volume of large and mid-size deciduous trees, density of mid-size trees, total woody stem density, total deciduous understory volume and total volume of large conifers were most frequently important in explaining variation in species abundances. Although each species had a unique set of structural affinities, multi-layered stands are apparently more important to long-distance migrants, in general, than to resident/short distance migrants. Large-scale efforts to identify important habitats, assess degree of protection, or monitor species/habitat trends are important to conservation. For forest birds, such efforts must include estimates of the factors to which the species of concern respond. At the stand scale in New England, it seems that bird abundance is more strongly associated with forest structure than with forest cover-type or stand size-class.

Effects of thinning and deer browsing on breeding birds in New England oak woodlands

Over a 3-year period we compared the effects of forest thinning and white-tailed deer (Odocoileus virginianus) browsing intensity on breeding birds in central Massachusetts oak (Quercus spp.) forests across 12 stands with dominant trees having a diameter at breast height greater than 27.9 cm. Unthinned stands with few deer had moderately dense woody understories and moderate ground cover; unthinned stands with many deer had sparse woody understories and little ground cover. Thinned stands with few deer had dense, tall woody understories and moderate ground cover; thinned stands with many deer had sparse woody understories and lush ground cover. Understory changes associated with high deer densities were apparently offset by effects of thinning so that thinned stands supported more breeding bird species. We recorded 65 species of birds; 28 occurred in all treatment classes. Thinned stands contained significantly (P < 0.05) more bird species than did unthinned stands. Of species that occurred more than once, nine occurred only in thinned stands, one occurred only in uncut stands, and another only in stands with 13–23 deer km−2. Occurrences of six species differed significantly (P < 0.05) by deer browsing intensity or thinning. Numbers of omnivores and ground gleaners (both species and individuals) were greater in thinned stands. Numbers of canopy gleaners were greater in stands with 1–3 deer km−2.

PREDATION ON REAL AND ARTIFICIAL NESTS IN SHRUBSTEPPE LANDSCAPES FRAGMENTED BY AGRICULTURE

Abstract Clearing of shrubsteppe communities for agriculture has created a highly fragmented landscape in eastern Washington, a condition that has been shown to adversely affect nesting success of birds in some forest and grassland communities. We used artificial nests monitored by cameras to examine relative effects of fragmentation, distance to edge, and vegetation cover on nest predation rates and to identify predators of shrubsteppe-nesting passerines and grouse. Predation rate for artificial nests was 26% (n = 118). Fragmentation had a strong influence on predation rates for artificial nests, with nests in fragmented landscapes about 9 times more likely to be depredated as those in continuous landscapes. Daily survival rate (± SE) for 207 real nests of 4 passerine species also was greater in continuous (0.978 ± 0.004) than in fragmented (0.962 ± 0.006) landscapes, although pattern of predation between real and artificial nests was not consistent among sites. Artificial nests were depredated by Common Ravens (Corvus corax), Black-billed Magpies (Pica hudsonia), Sage Thrashers (Oreoscoptes montanus), least chipmunks (Tamias minimus), and mice. Most nests in fragments were depredated by corvids (58%), whereas only Sage Thrashers and small mammals depredated nests in continuous landscapes. Increased predation by corvids and lower nest success in fragmented landscapes may have played a part in recent declines of some shrubsteppe birds. Future research should measure annual reproductive success of individual females and survival rates of juveniles and adults. Depredación de Nidos Naturales y Artificiales en Paisajes de Estepa Arbustiva Fragmentados por Agricultura Resumen. El reemplazo de estepa arbustiva por campos de cultivo ha creado un paisaje altamente fragmentado en el este de Washington, afectando adversamente el éxito de nidificación de aves en algunas comunidades de bosque y pastizal. Usamos nidos artificiales monitoreados por cámaras para examinar los efectos relativos de la fragmentación, la distancia al borde y la cobertura de la vegetación sobre las tasas de depredación de nidos, y para identificar los depredadores de paserinos y gallinas silvestres (Phasianidae) que nidifican en la estepa arbustiva. La tasa de depredación de los nidos artificiales fue del 26% (n = 118). La fragmentación tuvo una fuerte influencia en las tasas de depredación de nidos artificiales, ya que los nidos en paisajes fragmentados tuvieron una probabilidad de ser depredados 9 veces mayor que aquellos en paisajes continuos. La tasa de supervivencia diaria (± EE) de 207 nidos naturales pertenecientes a 4 especies de paserinos también fue mayor en paisajes continuos (0.978 ± 0.004) que fragmentados (0.962 ± 0.006), aunque el patrón de depredación entre nidos naturales y artificiales no fue consistente entre sitios. Los nidos artificiales fueron depredados por Corvus corax, Pica hudsonia, Oreoscoptes montanus, Tamias minimus y ratones. La mayoría de los nidos en fragmentos fueron depredados por C. corax (58%), mientras que sólo O. montanus y pequeños mamíferos depredaron nidos en paisajes continuos. Un incremento en la depredación por parte de C. corax y un menor éxito de los nidos en paisajes fragmentados puede haber jugado un rol en la disminución de algunas aves de la estepa arbustiva. Futuras investigaciones deberían medir el éxito reproductivo anual de hembras individuales y las tasas de supervivencia de juveniles y adultos.

Productivity of early successional shrubland birds in clearcuts and groupcuts in an eastern deciduous forest

Uneven-aged forest management has been advocated as a silvicultural practice because of concerns about the negative effects of even-aged management on birds that dwell in mature forests. Recent evidence, however, indicates that in the northeastern United States, bird species that inhabit early successional habitats may be experiencing more widespread declines than their mature-forest counterparts. We compared the effect of group selection, a widely used form of uneven-aged forest management, and clearcutting on nest survival rates of early successional shrubland birds in the White Mountains of New Hampshire. There was no difference in daily nest survival rate between clearcuts (0.990) and groupcuts (0.987) for 16 bird species combined (n = 290), and no difference in daily nest survival rate between clearcuts (0.993) and groupcuts (0.987) for chestnut-sided warblers (Dendroica pensylvanica), the only species for which enough nests were found for separate analysis (n = 217). There was no difference in daily nest survival rates of all species combined between edge (0.983) and interior areas (0.992) of clearcuts (n = 204), and no difference in daily nest survival rates of chestnut-sided warblers between edge (0.984) and interior (0.993) areas of groupcuts (n = 156). Thus, our results suggest that clearcuts and groupcuts provide similar habitat for species of early successional shrubland birds that inhabit both clearcuts and groupcuts. Recent studies, however, indicate that some bird species that use larger openings such as clearcuts do not occupy smaller openings created by group selection, which may limit the utility of group selection in managing habitat for early successional shrubland birds.

Nest predator distribution among clearcut forest, forest edge and forest interior in an extensively forested landscape

Abstract The distribution of avian and mammalian nest predator species in clearcut forest, forest edge and forest interior habitats was assessed on four study plots in 1992 and 1993 using line transect counts. Red squirrels (Tamiasciurus hudsonicus) and eastern chipmunks (Tamias striatus) were detected more often than expected by chance in mature forest than clearcut forest (P 0.05) and was not significantly correlated with conifer basal area or distance from clearcut borders (P>0.05). We conclude that the distribution of two important nest predator species, the eastern chipmunk and the red squirrel, are primarily influenced by the distribution of conifers. However, the distribution of eastern chipmunks and red squirrels are affected by clearcut borders as well.

Herpetofaunal species composition and relative abundance among three New England forest types

Drift fences and pitfall traps captured > 2000 reptiles and amphibians during 2 years; the most common species were wood frog (Rana sylvatica), American toad (Bufo americanus), and redback salamander (Plethodon cinereus). There were differences in species abundances among streamside and upland habitats within three forest cover types: northern hardwoods, red maple, and balsam fir. Among streamside stands, fewer (P < 0.05) individuals were captured in balsam fir. The two hardwood types supported the most species. Generally, more individuals were captured in upland than in streamside habitats. Both diversity (H′) and evenness (J′) were correlated with litter depth, and both were higher in hardwood than in balsam fir stands. Rana sylvatica, B. americanus and P. cinereus were more abundant in hardwood than in balsam-fir stands.