USDA Forest Service

Pacific Southwest Research Station


Pacific Southwest Research Station
800 Buchanan Street
West Annex Building
Albany, CA 94710-0011

(510) 559-6300

United States Department of Agriculture Forest Service. USDA logo which links to the department's national site. Forest Service logo which links to the agency's national site.

Research Topics Ecosystem Processes

^ Main Topic | Tropical Ecosystems | Sierra Nevada Ecosystems


Patterns of biodiversity, abundance and productivity of forest birds over an elevational gradient.

Study Plan
Research Project Summary

The Research

Forest Service lands include many forest types, broadly distributed along elevational and latitudinal gradients. Specific management activities are often concentrated in certain elevational bands and forest types, and wildlife species are affected according to their distribution and response. Identification of species' requirements for survival and reproduction and knowledge of their responses to forest management practices are crucial to maintaining biodiversity.

Our study focuses on estimating abundance and productivity of forest birds in four forest types across an elevational gradient. The four forest types include: ponderosa pine Censusing birds(1024-1372 m), mixed-conifer (1707-2012 m), true fir (2170-2347 m), and lodgepole pine (2469-2774 m). There are four replicate plots of each forest type, each consisting of 60 ha of mature forest with relatively high canopy cover. We censused birds on 8 (1995) or 16 (1996-2002) plots using a timed transect method. Transects are 1000 m long and are visited six times during the breeding season. Observers walk at a standardized rate, recording all birds seen or heard <50 and >50 m from the transect line.We search for nests and monitor them until they either fledge or fail. Open cup nests are checked directly, with a mirror on a pole, or a small video camera mounted on an extendable fiberglass pole. Cavity nests are checked with a fiberscope. Nests too high to be reached from the ground are climbed if the substrates are sturdy and healthy. We use a variety of climbing techniques to access nests and field assistants are trained and certified in Forest Service national tree climbing techniques. We collect habitat measurements at each nest and at random sites on each plot to compare nest sites and available habitat, and to compare successful and failed nests. We collect temperature data on each site and weather data are available from nearby weather stations.

Using fiberscope to examine cavity nestsResults to date show that species richness declines with increasing elevation. Abundance generally follows the same pattern, but abundance is highest in mixed conifer forests in mostyears. Cavity-nesters, however, show the reverse pattern, with most species attaining their highest abundance in true fir or lodgepole pine sites.

More than one third of the 96 species detected occur in all four forest types. Approximately one quarter of the species detected occurred in only one forest type. Of these, the vast majority (78%) occurred in ponderosa pine. Many of these species are typical of low-elevation, drier habitats, in vegetation types that are not well represented by public lands. These low-elevation forests provide important habitat for birds.

Abundance and productivity are variable across years, reinforcing the need for long-term Pacific-slope flycatcher neststudies. Many species had higher productivity at higher elevations. For two common species, dark-eyed juncos (Junco hyemalis) and dusky flycatchers (Empidonax oberholseri), the number of young fledged was negatively correlated with abundance, suggesting maladaptive habitat selection and that individuals are not able to judge the probability of nesting successfully in a particular habitat. There is some evidence of density-dependent predation for dusky flycatchers.

Tree species diversity also decreased with increasing elevation, with mixed conifer forests having the most tree species. Hardwood tree species, especially California black oak (Quercus kelloggii) and canyon live oak (Quercus chrysolepis) were used by a wide variety of species for nesting substrates and were used proportionately more than expected. In ponderosa pine forests, 45% of nests in live trees were in hardwoods. Although the importance of coniferous snags to wildlife cannot be overestimated, hardwood snags were seldom used by nesting birds.

Results of this study will also be valuable in examining potential responses by bird species to climate change, including changes in life history traits such as clutch size, number of young fledged, nest success, predation rate, timing of nesting, and the number of nesting attempts per season.

We hope to extend this study, using the mixed conifer forest sites, to study the effects of prescribed fire on birds. Due to heavy fuel loads present from fire suppression and risks to human habitation, prescribed burning is often limited to months of high moisture. In particular, prescribed fires initiated in the spring can coincide with nesting of breeding birds. Spring burning, its effects on neotropical migrant birds, and possible violations of the Migratory Bird Treaty Act have generated concerns over spring burning that require study. Baseline data provided by the current study will be valuable as annual variability can be great in these systems. We hope to study both short- and long-term effects and document changes in habitat structure following fire.


White-headed woodpecker nest
blue arrrowAssess patterns of abundance and productivity of bird species in four forest types over an elevational gradient.
blue arrrow Identify species' breeding habitat requirements, including identification of the most productive habitats for each species.
blue arrrow Assess the current population health of a wide range of species, including the identification of source and sink habitats.
blue arrrow Examine relations between abundance and productivity across the four forest types.
blue arrrow Develop models of habitat needs for healthy populations of coexisting species and to predict species' vulnerabilities to habitat change.
blue arrrow Examine elevational shifts in bird abundance, especially as they relate to the severity of the previous winter's weather, and to examine whether, in years of unusual weather conditions, variations in productivity are associated with these shifts.
blue arrrow Examine the importance of hardwood tree species to nesting birds
blue arrrow Predict the effects of climate change on forest birds, including changes in life history characteristics

Methods and Design

Eighteen study sites in four forest types were selected in 1994. Sites at the lowest elevations are in ponderosa pine stands (1025-1370 m elevation), followed by mixed-conifer stands (1710-2010 m elevation) and true fir stands (2170-2350 m elevation), with lodgepole pine stands at the highest elevations (2470-2775 m elevation). Each forest type has four replicates, except mixed conifer, which has six replicates. Of the six mixed-conifer sites, three lie in each watershed of the Kings River Project area. All sites will be protected from major disturbance for 10 years per a Memorandum of Understanding with the Kings River District of the Sierra National Forest, after which they will be incorporated into the landscape treatment for their watershed. Each site has a 40-ha gridded plot nested in approximately 60 ha of mature forest with relatively high canopy cover.

All plots are gridded at 50m intervals, with intersections marked with a steel fencepost and a cap of PVC pipe marked with an alphanumeric code to facilitate censusing and relocation of nests. Birds are censused on a 1000-m transect within the 40ha plot . Observers travel at a rate of 50 m per 3 minutes, recording all birds seen or heard. Three observers make two visits to each site each year, with each site counted six times per season. Only one observer will count a particular plot each day. The order of census visits and starting points of observers are randomly selected, with the constraint that visits are evenly divided between the two starting points. Censuses in the ponderosa pine forest type will be conducted between mid-April and mid-May, in the mixed-conifer type between the third week in May and mid-June, in true fire between late June and early July, and in lodgepole pine between late June and mid-July. Beginning census dates will be adjusted annually for each forest type, with the presence of singing birds the cue for the initiation of censusing. Censuses will begin at 0700 PDT in all forest types except the ponderosa pine type, where censuses will begin at 0730 PDT to accommodate the shorter day length earlier in the season.

All birds detected on each side of the transect line are recorded on the bird census form although detections before the start point or after the end point of the transect are not recorded. Detections of birds are recorded as <50 or as >50 m from the transect line to unlimited distance from the transect line, but within the same forest type. The location of the bird where it was first detected is recorded, avoiding double-counting individuals.

Nests of all bird species will be located on each plot and monitored. Nests will be checked every 3 to 4 days. Open nests on the ground or low in shrubs will be checked by simply looking into the nest. Higher nests will be checked with a mirror or small video camera mounted on an extendable pole. We will check cavity nests with a fiber optic instrument with a light source that allows us to see into dark and deep cavities. Using a variety of climbing techniques, we will climb to high nests in stable substrates once a week. On other nest visits and for nests that cannot be climbed, we will observe for activity, with the observation period lasting until activity is observed at the nest, or a minimum of 15 minutes if no activity is observed. Ideally, nests should be directly checked (if possible) at least once during incubation to obtain clutch size and identify nest parasitism by Brown-headed Cowbirds (Molothrus ater), close to the projected hatching date to determine date of hatching, immediately following hatching to determine the number of eggs hatched, and again 2 to 3 days before the expected fledging date to obtain an estimate of the number of young fledged. At nest checks, the minimum information needed is whether the nest is still active and the nesting stage ( i.e., laying, incubation, hatching, nestling). Nesting stage for nests that are observed only will be based on the behavior of the adults. Information on the age of nestlings is very helpful in determining nest fate, and information describing nestlings will be included in the instructions for the nest observation form. Two important features to note include whether the eyes are open and the degree of emergence of the flight feathers.

After a nest either fledges or fails, habitat measurements are taken at the nest site and at random sites on each plot, allowing identification of correlates of success and habitat selection. Data on ground and shrub cover will be recorded. The variables recorded to describe the nest site and the habitat surrounding the nest include, but are not limited to: basal area of trees and snags; number and species of trees and snags by size class; canopy cover; litter depth; slope; aspect; length, diameter, and decay class of logs, and percent cover of grasses, forbs, rock, soil, litter, logs, shrubs, and trees in the understory (< 1.3 m high). Variables specific to the nest include nest height; species, height, and diameter of the nesting substrate; and distance from the nest to the edge of the canopy and to the center of the nest substrate. For open nests, we will measure the number and average diameter of supporting branches, and concealment in the four cardinal directions, above, and below the nest. For cavity nests, we will measure vertical and horizontal entrance diameters, depth, bottom area, orientation of the opening, and entrance length. Collection of nest site data will often involve climbing to the nest. Data that cannot be measured directly will not be estimated.

To characterize available habitat, vegetation measurements consisting of a subset of those used at nests will be done at ten to fifteen randomly selected grid points on each plot each year. A random distance and direction from the alphanumeric grid marker will be chosen from random number tables, with the stipulations that the locations of all points are 100 m apart and at least 100 m from site boundaries. Information on the nearest snag, tree, sapling, shrub, and stump within 35 m will be recorded.

Annual temperature and precipitation data will be obtained from local weather stations, to index the severity of the winter preceding the breeding season. Hourly temperature readings will be recorded at two sites in each forest type during the breeding season using temperature data loggers. Data loggers will be placed inside a white box mounted to a fence post, approximately 1 m above ground, and located in a clearing on the plot having a minimum diameter of 30 m. During years when snow remains on the plots during censuses, we will record the percent cover and depth of snow in a 1 m radius circle around each grid marker on the census line.

Nest fates will be determined based on information on the nest observation form. We will use our own nest data as it accumulates to determine the length of the incubation and nestling periods. Nests that fledge at least one young will be considered successful. Nest success and daily mortality rates will be calculated using the logistic exposure method.

Using species-specific survival rates from the literature and habitat-specific productivity from this study, I will develop population models to identify source and sink habitats. The relationship between abundance and productivity across forest types will be examined using correlation analysis. I will use multiple regression to examine the relations between abundance and habitat and nest-site variables and generalized linear models to examine relations between nest success and habitat and nest-site variables, using AIC to rank candidate models.

Application of Research Results

Forest Service lands are broadly distributed along elevational and latitudinal gradients, with specific management activities often concentrated in certain elevational bands and forest types.. Wildlife species are affected according to their distribution and response. To maintain biodiversity, we need identify species' requirements for survival and reproduction and predict their responses to forest management practices.

Results obtained from this research are important to Forest Service resource specialists and biologists from other agencies and nongovernmental organizations in managing for and maintaining biodiversity. For example, California Partners in Flight has requested information from the PI for developing habitat conservation plans to ensure the long-term conservation of birds and their habitats within California.

Identification of source and sink habitats and an understanding of their dynamics are crucial for maintaining healthy populations of forest birds. Our models predicting species' responses to severe or unusual weather conditions will enable biologists to anticipate the range of natural variability in abundance and productivity of forest birds, to understand the dynamics of movements and habitat selection between forest types, and to foresee the results of these movements.


map of the Sierra Nevada Ecoregion
This study is being done in the High Sierra Ranger District of the Sierra National Forest in conjunction with the Kings River Administrative Study.

Lead Scientists/Collaborators

1.)Purcell, K, and 2.)Drynan, D..

USDA Forest Service
Pacific Southwest Research Station
Sierra Nevada Research Center
2081 E. Sierra Avenue
Fresno, CA 93710

1.)ph: 559-868-6233
fax: 559-868-3491

2.)ph: 559-323-3208
fax: 559-297-3355

Publications and Reports

Purcell, K. L. 1997. Use of a fiberscope for examining cavity nests. Journal of Field Ornithology 68:283-286.

Purcell, K.L., and J. Verner. 1999. Abundance and rates of brood parasitism by Brown-headed Cowbirds over an elevational gradient in the southern Sierra Nevada. Studies in Avian Biology 18:97-103.

Purcell, K.L. 2002. Abundance and productivity of birds over an elevational gradient. Pages 121-132 in J. Verner (technical editor). Proceedings of the Symposium on the Kings River Sustainable Forest Ecosystems Project: Progress and Current Status. USDA Forest Service Gen. Tech. Rep. PSW-GTR-183. Albany, CA.

North, M., B. Oakley, J. Chen, H. Erickson, A. Gray, A. Izzo, D. Johnson, S. Ma, M. Mara, M. Meyer, K. Purcell, T. Rambo, D. Rizzo, B. Roath, T. Schowalter. 2002. Vegetation and ecological characteristics of mixed-conifer and red fire forests at the Teakettle Experimental Forest. USDA Forest Service Gen. Tech. Rep. PSW-GTR-186. Albany, CA. 52 pp.

Last Modified: Apr 19, 2011 08:03:28 PM