William C. McComb

IDENTIFYING GAPS IN CONSERVATION NETWORKS: OF INDICATORS AND UNCERTAINTY IN GEOGRAPHIC‐BASED ANALYSES

Mapping of biodiversity elements to expose gaps in conservation networks has become a common strategy in nature-reserve design. We review a set of critical assumptions and issues that influence the interpretation and implementation of gap analysis, including: (1) the assumption that a subset of taxa can be used to indicate overall diversity patterns, and (2) the impact of uncertainty and error propagation in reserve design. We focus our review on species diversity patterns and use data from peer-reviewed literature or extant state-level databases to test specific predictions implied by these assumptions. Support for the biodiversity indicator assumption was varied. Patterns of diversity as reflected in species counts, coincidence of hot spots, and representativeness were not generally concordant among different taxa, with the degree of concordance depending on the measure of diversity used, the taxa examined, and the scale of analysis. Simulated errors in predicting the occurrence of individual species indicated that substantial differences in reserve-boundary recommendations could occur when uncertainty is incorporated into the analysis. Furthermore, focusing exclusively on vegetation and species distribution patterns in conservation planning will contribute to reserve-design uncertainty unless the processes behind the patterns are understood. To deal with these issues, reserve planners should base reserve design on the best available, albeit incomplete, data; should attempt to define those ecological circumstances when the indicator assumption is defensible; should incorporate uncertainty explicitly in mapped displays of biodiversity elements; and should simultaneously consider pattern and process in reserve-design problems.

Bird Habitat Relationships in Natural and Managed Forests in the West Cascades of Oregon

Ecologists have advocated retaining various densities of canopy trees in harvest units in Pacific Northwest forests. In contrast to clear-cutting, this practice may better emulate the patterns of disturbance and structural complexity typical of natural forests in the region. Several ecological attributes, including vertebrate habitat diversity, are thought to be associated with stands of complex structure. The goal of this study was to determine bird abundance in canopy retention sites relative to other common stand types in the Pacific Northwest and to develop habitat functions for extrapolating bird abundance across current and future landscapes. We used data from five previous studies in the west central Cascades of Oregon to compare bird abundance and to develop habitat functions for forest birds across a wide range of natural and managed stand structures and ages. The 67 stands included clearcuts, retention sites, young closed-canopy plantations, mature stands, and old-growth stands. ANOVA revealed that 18 of the 23 species included in the analysis differed significantly in abundance among the stand types, with some species being primarily associated with each of the stand types. The habitat variables used to build habitat functions included tree density by size class, mean tree diameter, and variation in tree diameter. Linear, polynomial, and various nonlinear regression models were evaluated for each bird species. Significant habitat functions were generated for 17 of the 23 bird species. The analyses identified four habitat-use guilds among the 17 bird species: open-canopy; open-canopy with dispersed large trees; structurally complex closed-canopy; and structurally simple closed-canopy guilds. This study is the first in the Pacific Northwest to compare bird abundances across natural stands, traditionally managed plantations, and stands managed under ecological forestry approaches. The results suggested that canopy tree retention benefits many, but not all, of the bird species we studied, Moreover, the nonlinear responses of bird abundance revealed thresholds in tree density at which bird abundance changed dramatically. Knowledge of these thresholds allow managers to design stands for specific biodiversity objectives. The habitat functions presented here can be used to predict bird abundance based on habitat measurements derived from field data, remotely sensed data, or output from computer models of forest dynamics.

Alternative Silvicultural Regimes in the Pacific Northwest: Simulations of Ecological and Economic Effects

New silvicultural strategies to sustain both ecological and human communities are being developed and implemented on federal forest lands in the Pacific Northwest (PNW) United States. Two important stand-level components of the new silviculture regimes are rotation age and retention level of live trees in harvest units. Ecologists have suggested that canopy tree retention and longer rotations will create patterns of stand structure in managed forest that are similar to those in natural forests, and promote long-term ecological productivity and biodiversity. Forest economists, however, are concerned that canopy tree retention and long rotations may reduce wood production, although the high value of large logs produced by these new silvicultural regimes may compensate for reduced growth rates. We used the forest model ZELIG to perform a factorial simulation experiment on long-term responses of ecological and economic variables to nine retention levels and four rotation lengths. ZELIG output on forest structure and composition was input to a forest economics model that calculated net value of wood products in 1989 dollars. The simulated stand data were also linked with regression equations to predict the densities of 17 bird species as a function of tree size class distribution. Five replicates of each treatment were run for the 240-yr simulation period. Results indicated that stand structure under each of the canopy tree retention levels was more similar to the pre-treatment natural forest than following clear-cutting. Variation in tree size under intermediate levels of retention, however, did not reach the level of the natural forest during the simulation period. Tree species composition was strongly related to retention level and rotation age. Shade-intolerant Douglas- fir (Pseudotsuga menziesii) lost dominance to shade-tolerant species under intermediate retention levels and longer rotations. Wood production decreased significantly with increasing retention level and rotation age, with a notable threshold between retention levels of 0 and 5 trees per hectare. Net wood products value did not decrease as rapidly with retention level, and did not differ much among rotation ages, because of the high value of large logs. Bird species responded individualistically to retention level and rotation age. Some had peak densities under short-rotation clear-cutting, but most were associated with structurally complex, closed-canopy forest. Consequently, bird species richness increased significantly with retention level and rotation age. Within the assumptions and limitations of our models, this application provided knowledge on trends and thresholds that can help land managers to choose silvicultural regimes that best balance their management objectives. We concluded that retention level and rotation age strongly influence ecological and economic responses in PNW forests; efforts are needed to reduce uncertainty about these effects.

BREEDING BIRD RESPONSES TO THREE SILVICULTURAL TREATMENTS IN THE OREGON COAST RANGE

Silvicultural alternatives to clear-cutting have been suggested to promote development, retention, or creation of late-successional features such as large trees, mul- tilayered canopies, snags, and logs. We assessed bird response to three silvicultural alter- natives to clear-cutting that retained structural features found in old Douglas-fir (Pseudo- tsuga menziesii) forests and that imitated natural disturbance regimes more closely than did traditional clear-cutting: (1) small-patch group selection treatment representing a low- intensity disturbance; (2) two-story treatment, representing a moderate to high-intensity disturbance; and (3) modified clear-cut treatment, representing a high-intensity disturbance. We counted diurnal breeding birds 1 yr prior to and 2 yr after harvest to estimate effects of the silvicultural treatments on bird communities compared with uncut controls. The small-patch group selection treatment was most similar in species composition to control stands. The two-story treatment was more similar to the modified clear-cut treatment. Ten bird species remained abundant following the small-patch group selection treatment. They declined in abundance in modified clearcuts and two-story stands. These species included four neotropical migratory species and five species with restricted geographic ranges and habitat associations. Nine species increased in response to moderate and/or high-intensity disturbances. This group included a larger proportion of species that were habitat generalists. Silvicultural treatments imitating low-intensity disturbances were most effective in retaining bird communities associated with mature forest; high-intensity disturbances such as the two-story and modified clear-cut treatments greatly altered bird community composition. Bird responses to the silvicultural treatments that we studied indicate that a variety of stand types is needed to meet needs of all species.

Microclimates of Nest Boxes and Natural Cavities in Bottomland Hardwoods

BELLROSE, F. C. 1976. Ducks, geese, and swans of North America. Stackpole Books, Harrisburg, Pa. 543pp. JONES, R. E., AND A. S. LEOPOLD. 1967. Nesting interference in a dense population of wood ducks. J. Wildl. Manage. 31:221-228. KLEIN, H. G. 1955. Wood duck production and use of nest boxes on small marshes in New York. N.Y. Fish and Game J. 2:68-83. LUCKETT, L. M. 1977. The ecology of wood duck (Aix sponsa) in the Piedmont region of South Carolina. M.S. Thesis. Clemson Univ., Clemson, S.C. 99pp. MCGILVREY, F. G., AND F. M. UHLER. 1971. A starling-deterrent wood duck nest box. J. Wildl. Manage. 35:793-797. NORMAN, F. I., AND T. L. RIGGERT. 1977. Nest boxes as nest sites for Australian waterfowl. J. Wildl. Manage. 41:643-649.

Selection of Day Roosts by Female Long-Legged Myotis in the Central Oregon Cascade Range

We radiotracked 16 female long-legged myotis (Myotis volans) in the central Oregon Cascade Range that used a total of 41 day roosts. Large Douglas-fir (Pseudotsuga menziesii) snags averaging 97 ± 7 (SE) cm diameter at breast height (dbh) and 38 ± 3 m in height were the most commonly used roost structures (88%). The odds that a snag was used as a day roost increased as snag height increased (P < 0.001); after snag height was accounted for, the odds of use decreased as stand height within 20 m of the snag increased (P = 0.024). The frequency of occurrence of roosts between young and late seral stands did not differ from that by chance in these 2 stand conditions (P = 0.76). Day roosts generally occurred in upland habitats associated with streams that contained night roosts. Management of large diameter, tall snags that extend above the canopy will provide 1 component of day-roost habitat for long-legged myotis in managed landscapes.

Response of Small Mammals to Clearcutting, Burning, and Glyphosate Application in the Oregon Coast Range

We sampled amphibians on 3 red alder (Alnus rubra) sites 1 year before and 1 and 2 years after the following treatments were applied to each site: (1) control (uncut), (2) clearcut and broadcast burned, and (3) clearcut, broadcast burned, and then sprayed with the herbicide glyphosate. All sites included uncut riparian buffer strips. For 3 of the 6 species with ≥20 captures in pitfall traps, we did not detect changes in capture rates after clearcutting. Capture rates of ensatinas (Ensatina eschscholtzii) and Pacific giant salamanders (Dicamptodon tenebrosus) decreased after logging. Capture rates of western redback salamanders (Plethodon vehiculum) increased the first year after logging, probably because the salamanders sheltered in pitfalls, but effects on populations were unclear. Logging did not significantly alter capture rates of rough-skin newts (Taricha granulosa), Dunns salamanders (P. dunni), and red-legged frogs (Rana aurora). Planning the location and timing of clearcuts or other silvicultural practices over a landscape and retaining riparian buffer strips may be necessary to ensure long-term persistence of Pacific giant salamanders. We did not detect any effects of herbicide spraying on capture rates. Capture rates for rough-skin newts and red-legged frogs were higher in uncut red alder stands than in Douglas-fir (Pseudotsuga menziesii) stands sampled in other studies, an indication that, when red alder stands are converted to Douglas-fir, some alders should be left adjacent to streams to provide habitat for these species and other hardwood associates.

Effects of silvicultural modifications of temperate rainforest on breeding and wintering bird communities, Prince of Wales Island, southeast Alaska

We inventoried breeding and wintering bird communities in four treatments of temperate rainforest on Prince of Wales Island, southeast Alaska during 1991-1992 and 1992-1993. The four forest treatments sampled included: (1) young growth (20 years) originating from clearcut logging with no silvicultural modification (non-modified), (2) young growth (20 years) precommercially thinned along uniformly-spaced thinning grids (thinned), (3) young growth (20 years) with gaps in the overstory canopy created by felling trees in 0.05-ha openings (gapped), and (4) virgin old growth (≥ 150 years). Of 16 common breeding bird species observed, six showed significant responses to young-growth modifications. One species was more abundant and two species were less abundant in thinned sites, while one species was more abundant and two species were less abundant in gapped sites than at least one of the other treatments. None of the three common wintering species of birds observed was influenced by young-growth modification. Breeding bird communities, in general, were less similar between young- and old-growth treatments than among young-growth treatments. Three of the 16 common breeding bird species were more abundant in old growth than each of the young-growth treatments and one uncommon species was detected almost exclusively in old growth during both the breeding and wintering seasons. Four other breeding bird species were more abundant in young-growth treatments than in old growth. Higher use of old growth by wintering birds was related to winter severity. To enhance habitat for wintering and breeding birds we recommend: (I) thinning young growth along variable-spaced grids to create additional canopy layers and improve snow-intercept properties of young growth for canopy-foraging birds, (2) retention of old-growth clumps in clearcuts for bird species associated with old-growth structure, and (3) long-term conservation of old-growth temperate rainforest for breeding and wintering birds positively associated with old growth.

Nest-Box and Natural-Cavity Use in Three Mid-South Forest Habitats

Use of 235 nesting boxes by birds and mammals was compared to use of 165 natural cavities in 3 forest habitats. Nest boxes were used more frequently than natural cavities by 12 of the 19 species using cavities (P < 0.05). Average use of nest boxes after 2 years was constant regardless of nest-box size, density of nest boxes, or habitat type. Most birds and mammals appeared to select cavities on the basis of 1 or more physical characteristics of the cavity and/or the surrounding area. Nest boxes or natural cavities having desirable characteristics and located in residual trees after logging (thinning or harvesting) may increase seasonal use of each cavity. Interspersion of nest boxes among natural cavities may aid species dependent upon natural cavities by reducing interspecific competition for natural cavities. J. WILDL. MANAGE. 45(1):93-101 Bottomland hardwoods are being cleared and replaced with agricultural crops and hardwood plantations. Sternitzke (1976) suggested that if the midSouth is to continue as a leading manufacturer of hardwood products, more intensive silvicultural practices must be implemented in upland hardwoods. Several authors have reported that intensive silvicultural practices adversely affect cavity-dependent wildlife (Thomas et al. 1976, Hardin and Evans 1977, DeGraaf 1978, Scott 1979). Seemingly, the only practical method of providing nesting and denning sites for cavity-dependent wildlife, while managing hardwood forests for maximum cellulose production, is to use nest boxes. The objectives of our study were to compare use of nest boxes with use of natural cavities in an eastern cottonwood (Populus deltoides) plantation, in an upland hardwood-pine stand, and in a bottomland hardwood stand; and to determine which physical characteristics of the cavity and the surrounding area were important to species using cavities. Funding was provided by the Louisiana Agricultural Experiment Station, McIntire-Stennis Project 1893. We thank P. E. Schilling and J. L. Stewart for statistical help; and H. R. Perry, Jr., J. C. Chambers, and T. Hansbrough for aid and advice.

Cumulative effects of forestry practices: an example framework for evaluation from Oregon, U.S.A

Abstract Cumulative effects of forestry operations accumulate over time and space in forested landscapes where harvesting and management occur. We review the literature and concepts associated with cumulative effects and propose a framework for evaluating them. In order to evaluate potential adverse effects of forestry on vegetation, soils, streams, aquatic organisms, wildlife and air, baseline conditions and natural variations of resource characteristics must be known. Cause-effect relationships must be documented. Systems of measurements and monitoring must be implemented along with databases and geographic information systems for displaying information at spatial scales from individual sites to landscapes and regions. Systems for decision-making must be implemented. We provide an example of a framework for such a system in a mountainous, forested river basin in northwest Oregon, U.S.A. We conclude that knowledge and technologies for preliminary systems exist now, but that for more refined systems more knowledge of details of cause-effect links and of simulation models should be developed.