Conservation of special status species includes identifying and focusing on primary habitats, the environmental conditions that are tied to their highest likelihood of occurrence, rather than secondary habitats where a species occurs at lower density (Goble et al. 2012). This process allows land managers and conservationists to prioritize specific locations for protection where maximum benefits can be achieved with limited conservation resources. In some cases secondary habitats may also be important for long-term population viability (Howe et al. 1991, Aldridge and Boyce 2007). Secondary habitat can offer refugia especially if primary habitat is reduced in quantity or quality (Nielsen et al. 2006). However, because many species of conservation concern are rare even in primary habitat, identifying and managing secondary habitat can be a challenge.
Regional monitoring programs provide data for multiple species and create opportunities to inform species conservation decisions where otherwise no data would have been available (Geupel et al. 2011). Though such monitoring programs may not be specifically designed for evaluating population abundance and distribution for certain species, modern occupancy modeling methods have been developed that account for imperfect detectability, thus making it possible to reliably estimate species distributions and habitat associations even in some cases for rare and inconspicuous species (MacKenzie et al. 2006). Occupancy modeling can also account and correct for detectability.
In western North America, the challenge of managing primary and secondary habitats to conserve a bird population is exemplified by the Black-backed Woodpecker (Picoides arcticus). The Black-backed Woodpecker is considered a habitat specialist. Throughout its range, it occurs at highest densities in conifer forests that have recently burned (within 1-5 years) at high severity (Murphy and Lehnhausen 1998, Hutto 2008, Saracco et al. 2011). In western North America, especially in the Sierra Nevada and Rocky Mountain ranges, research and management has primarily focused on burned forest habitat (Dixon and Saab 2000, U.S. Forest Service 2007, Hutto 2008, Saracco et al. 2011, Bond et al. 2012). Their status and occurrence patterns in Sierra Nevada green forest are unknown, although small numbers of nesting pairs have been documented outside of areas that recently burned at moderate or high severity (Raphael and White 1984). The importance of this secondary habitat may be underappreciated by researchers and managers working to conserve this rare species at the southern extent of its range (Odion and Hanson 2013).
In the absence of a regional population estimate and a clear understanding of the relative importance of burned forest (primary habitat) and green forest (secondary habitat), there is increasing debate regarding the level of protection Black-backed Woodpeckers should receive (Hanson and Cummings 2010, Hanson et al. 2012). The species was recently considered for listing as threatened or endangered in California (California Department of Fish and Game 2012) and the California population is currently being considered for listing under the United States Endangered Species Act (Federal Register 2013).
In the Sierra Nevada occupancy in burned forest approaches 0.60; with strong positive associations with elevation, northern latitudes, and high fire severities that produce high snag density (Siegel et al. 2012). Occupancy in green forest is unknown, but many assume it to be near zero (Hanson and Cummings 2010, Hanson et al. 2012). Given that there is vastly more green forest than burned forest, even a low density population of Black-backed Woodpeckers in green forest may be important to conservation of this species in the region.
Understanding the habitat associations of Black-backed Woodpecker in green forest is an important step to developing management guidelines for the species. Their habitat associations in green forest of western North America, and especially the Sierra Nevada are poorly known. In eastern Canadian boreal forest, they are associated with late seral characteristics including large amounts of downed wood and snags (Tremblay et al. 2009), and large diameter trees and snags (Hoyt and Hannon 2002, Tremblay et al. 2010). Similarly, in the Cascade Mountains of Oregon, Black-backed Woodpeckers nested in mature and over-mature lodgepole pine stands with an abundance of snags and beetle-infested trees (Goggans et al. 1989). They have also nested in forests with mountain pine beetle (Dendroctonus ponderosae) outbreaks in the Black Hills of South Dakota (Bonnot et al. 2009).
Beyond the importance of conserving a single species, protection of Black-backed Woodpeckers may have broader implications to the cavity nesting wildlife community in North American conifer forests. Woodpeckers are often considered keystone species for their role in excavating cavities used by many other wildlife species (Martin and Eadie 1999, Virkkala 2006, Tarbill 2010). Black-backed Woodpeckers may play a unique role among primary cavity excavators in that they are known to excavate cavities in less decayed snags than other woodpecker species (Raphael and White 1984, Seavy et al. 2012) and rarely reuse cavities from year to year (Saab et al. 2004). Thus, they are able to produce cavities that may be used by other species for several decades in forest where existing decayed snags are lacking. This may be important for allowing cavity nesting species to exploit ephemeral food resources, such as bark and wood-boring beetle outbreaks (e.g., Norris and Martin 2010). It has been suggested they play a keystone role for the cavity nesting wildlife community in eastern boreal forest (Tremblay et al. 2009, 2010), and burned forest of the Sierra Nevada (Tarbill 2010).
To begin to address the knowledge gap in Black-backed Woodpecker distribution and habitat requirements in western North American forests and help guide management of this potential keystone species, we used data from a Sierra Nevada avian monitoring program to investigate their occurrence in green forests of the Sierra Nevada. Using an occupancy modeling framework, our objectives were to (1) identify the important habitat types and attributes to guide management of this species across its range in California, (2) use multiyear dynamic models to estimate occupancy in green forest, (3) use extinction and colonization probabilities to infer the persistence of birds on occupied sites, and (4) summarize information on Black-backed Woodpecker nests encountered in Sierra Nevada green forest.
Our study area included nine national forests and the Lake Tahoe Basin Management Unit in the Sierra Nevada forest planning area in California, USA (Fig. 1; U.S. Forest Service 2004a). This area extends from the southern Cascade and Warner Mountains in Modoc National Forest to Sequoia National Forest at the southern extreme of Black-backed Woodpecker range. Maximum elevations across this region are significantly higher in the southern Sierra vs. the northern Sierra and southern Cascade Mountains (4200 m in the south vs. 2900 m in the north). Precipitation patterns are Mediterranean with the vast majority occurring from November to March. Precipitation increases with elevation and substantially higher amounts fall on the western slope compared with the rain-shadowed eastern slope. Conifer forest is the dominant land cover with ponderosa pine (Pinus ponderosa) types dominant at lower elevations; mixed conifer forests comprising ponderosa pine, white fir (Abies concolor), sugar pine (P. lambertiana), Douglas fir (Pseudotsuga menzeisii), and incense cedar (Calocedrus decurrens) are dominant at intermediate elevations; and white fir, red fir (A. magnifica), Jeffrey pine (P. jeffreyi), and lodgepole pine (P. contorta) are dominant at higher elevations.
Before ~1940, fire was common in the Sierra Nevada with short fire return intervals (5-20 years) in mixed conifer and lower elevation east and west-side pine-dominated forests; longer intervals (15-60 years) occurred in higher elevation red fir and lodgepole pine forest (Stephens et al. 2007). Since ~1940, fire suppression efforts have led to a significant decline in area burned annually across the Sierra Nevada, but recent evidence has shown that when fires currently burn, they are larger and burn at higher severity than in the beginning of the previous century (Miller et al. 2009a). A sample of the California Fire Return Interval Departure GIS layer (Safford and Van de Water 2013, U.S. Forest Service 2013) at our survey locations show that approximately 21% of the study area has burned in the last 70 years, and on average each location has gone more than four fire return intervals without burning. High-elevation red fir and lodgepole pine forests have a lower departure from their historic (pre-Euro-American settlement) fire return interval than low and middle-elevation mixed conifer and pine-dominated forest (Mallek et al. 2013). Thus, they may be closer to their presettlement forest structure.
We recorded all Black-backed Woodpeckers detected at sampling locations established for a bioregional monitoring project that was not originally designed for monitoring this species, but rather for four other avian forest and chaparral management indicator species (Roberts et al. 2011). Sample locations ranged in elevation from 1005 to 2737 m and latitudes from 35.3929° to 41.9401°, and were limited to areas within 1 km of accessible roads or trails and slopes less than 35%. We assembled a vegetation layer of 35 different California Wildlife Habitat Relationship (CWHR) land cover types (U.S. Forest Service 2004b). We then eliminated nonforest and nonmontane habitat types including grassland, sagebrush, riparian, foothill chaparral, oak woodland, subalpine, barren, and juniper. These stratifications reduced our sampling frame to 1,505,500 ha.
Sample locations were selected using a generalized random-tessellation stratified (GRTS) sampling protocol (Stevens and Olsen 2004, Theobald et al. 2007). The set of potential sampling locations was built from a tessellation generated in ArcGIS (Ver. 9.2, Environmental Systems Research Institute, Redlands, California, USA), consisting of a grid of 1 km² cells with a random origin covering the entire study area. We used the GRTS algorithm to select survey locations with equal weight across the entire study area, which consisted of 13,976 1 km² cells, resulting in the placement of locations proportional to the amount and spatial distribution of all forested and chaparral habitats within our sampling frame. At each location we established two transects in adjacent 1 km grid cells consisting of four point count stations at 250 m in the cardinal directions from a fifth point in the center. This resulted in a sample of 2300 points on 460 transects distributed as 230 spatially balanced pairs.
To evaluate occurrence patterns in green forest we removed sampling locations associated with their primary burned forest habitat. We used Spatial Analyst in ArcGIS (Ver. 10.1, Environmental Systems Research Institute, Redlands, California, USA) to calculate the distance from each transect to the nearest patch burned at moderate or high severity within the previous eight years for each of the three years we sampled for Black-backed Woodpecker (2011-2013). We accessed composite burn index fire severity information from the California Vegetation Burn Severity GIS layer (U.S. Forest Service 2012) and used the definitions of moderate and high severity described by Miller et al. (2009b). We removed transects where the average distance to the nearest moderate or high severity fire patch for the five point count stations was less than 2 km. We also removed all transects with at least one point occurring directly within a patch that had burned at moderate or high severity since 1990. We included locations within areas that had burned more than 20 years prior since that is well beyond the time that Black-backed Woodpeckers are known to occupy those habitats (Dixon and Saab 2000, Saracco et al. 2011). We feel this definition gives us a conservative sample of the Black-backed Woodpecker population in green forest. Any individuals detected more than 2 km from recently burned patches or within older fire areas are best classified as green forest birds. Removing these locations reduced our sample from 460 transects to the 386 we have defined as green forest.
At each of the five point count stations within a transect we conducted a standardized unlimited distance 5 min point count survey (Ralph et al. 1995). At each station a single observer estimated the distance to the location of each individual bird they detected (hereafter “passive surveys”). Following the five passive surveys, at the center point of each transect only, we conducted a 5 min playback survey for Hairy Woodpecker (Picoides villosus) and Mountain Quail (Oreortyx pictus), and a 6 min playback survey for Black-backed Woodpecker. We conducted surveys for the two other species as part of the protocol for the regional monitoring program for which these sampling locations were established. Black-backed Woodpecker survey duration was 6 min, with a series of three 25 sec playbacks followed by 95 sec of listening and watching. Playbacks included the scream-rattle-snarl and pik calls and territorial drumming sounds (recording by G. A. Keller, Macaulay Library of Natural Sounds, Cornell Laboratory of Ornithology). Playbacks were broadcast at a standardized volume (90 db) using FOXPRO® ZR2 digital game callers (FOXPRO Inc., Lewistown, Pennsylvania, USA). Playback surveys have been shown to significantly increase detection probability for this species compared to individual passive point count surveys (Saracco et al. 2011). Playback surveys were only conducted once per transect visit after all passive point count surveys were completed to avoid influencing detection probability on passive surveys via individuals drawn toward the broadcast from large distances away. Based on our field observations, the approximate range at which human observers can hear the playback calls is 200 m, but highly variable depending on topography and vegetation. There were three transects where the only Black-backed Woodpecker detections were during a Hairy Woodpecker/Mountain Quail playback survey, and none were detected during the passive surveys or Black-backed Woodpecker playbacks. Because of our relatively small sample of Black-backed Woodpecker detections, we included those data in our analysis to better inform our models.
All observers underwent an intensive, three-week training period focused on bird identification prior to conducting surveys. Surveys were conducted between local sunrise and 1000 h from 13 May to 15 July in 2011 to 2013. Surveys did not occur in inclement weather that could reduce detectability, e.g., high wind, rain, or dense fog. In each year, at least 96% of transects were surveyed, and 93% were visited in all three years. The remaining 7% were visited in two out of the three years. Of the transects visited in all three years, 48% received two visits in 2011, 80% in 2012, and 62% in 2013, with the remainder receiving a single visit. Variable survey effort was accounted for in our occupancy modeling framework described below.
We assigned a CWHR habitat type to each of the point count locations using a GIS framework (Mayer and Laudenslayer 1988, U.S. Forest Service 2004b), and then selected the CWHR type representing the largest number of points in each transect. To reduce the number of predictor variables in the occupancy model we combined the full list of 14 different CWHR types into the following six categories with similar vegetation structure and composition; Sierra mixed conifer (includes ponderosa pine, Douglas fir, and pine-hardwood types), eastside pine (includes Jeffrey pine), white fir, red fir, lodgepole pine, and all remaining nonforest types, including grassland, chaparral, juniper, and sagebrush, were combined into “other.” We used a statewide digital elevation model in GIS to assess elevation (as a residual of regression with latitude to remove correlation), percent slope, and solar radiation index (SRI), which is a linear representation of aspect (Keating et al. 2007).
At each point we quantified forest structural conditions within a 50 m radius plot (0.79 ha) centered at each point count station. We made ocular estimates of percent cover of trees > 5 m in height, measured basal area using a 10 factor Basal Area Factor key gauge (converted to per hectare scale), average diameter of the largest trees in the plot, and counted number of snags >10 cm in diameter. To create transect-scale habitat covariates for occupancy models we averaged the tree cover, basal area, and largest tree diameter at breast height (dbh) values from the five points on each transect. To characterize snag resources, we calculated the maximum snag count from among the five points rather than average. We used the maximum snag count because woodpeckers are readily able to travel the 250 meters between points and thus averaging across the five points would not describe these resources as the birds might perceive them.
To document breeding activity by Black-backed Woodpeckers, we recorded information for 19 nests found in green forest. Nests were found incidentally by trained field technicians during the May to July survey period during avian point count monitoring in the Sierra Nevada from 2004 to 2013. In 2012 we instructed field crews to follow any individual they observed to find nests as time allowed. In all other years, nests were from chance encounters with far greater survey effort in the northern Sierra. For each nest, we recorded the nest tree condition (snag or live tree), nest tree species, CWHR habitat type, and for nine nests we recorded tree or snag dbh.
To evaluate the distribution and habitat associations of Black-backed Woodpecker in Sierra Nevada green forest we used occupancy models (MacKenzie et al. 2006). Occupancy models correct occurrence data for imperfections in the detection process, variable sampling efforts across time and space, and can include covariates on both probability of detection and occurrence by assessing the history of detections across multiple visits to the same survey sites.
We assembled detection histories for each transect by combining all detections from the five passive point counts during a single transect visit, and considered this as a separate survey event from the playback surveys that were done only at the center point. We visited each transect up to twice per year, for a maximum of K = 4 survey events (2 passive, 2 playback) per year per transect. For all models we included survey type (passive or playback) as a covariate of detection probability.
To answer the questions posed, and because of the relatively small sample of detections, we built three separate occupancy models. First we used two single-season models to determine occupancy associations by (1) forest type and (2) habitat structure and physiography. We separated these into separate analyses because of the correlation between habitat types and the physiographic variables. We estimated occupancy and site turnover rates across the three years of our study period with a multiseason dynamic occupancy model.
To evaluate forest type and habitat associations we used a static single-season occupancy model where each site was defined as a transect-year combination, thus “stacking” the yearly detection histories at all surveyed transects to achieve a larger effective sample size (hereafter referred to as the “static habitat association model”). By using this design, we were able to include a larger number of habitat covariates for model selection and avoid problems associated with poor model fit experienced with our full set of covariates using the multiseason occupancy model. The total sample size for this dataset is n = 1135 sites (transect/year combinations) with 377 transects visited in 2011, 386 in 2012, and 372 in 2013. This effectively triples the sample size for examining the relationships between environmental variables and occupancy. Replicating the sites for each survey year in this manner does have the potential to underestimate error in model coefficients due to pseudo-replication, but we feel this method is preferable to the multiseason dynamic model because sites with potentially higher quality habitat are likely to be occupied across more years than lower quality sites and thus will contribute more to identifying clear habitat associations with these sparse data (McClure and Hill 2012). In the multiyear dynamic formulation a site that is occupied only one year, i.e., a potentially lower quality habitat, will have as much influence on the occupancy-habitat relationship as sites that are occupied consistently across years.
In the static habitat association model, we included a large set of covariates on occupancy including; latitude (degrees), the residual of elevation regressed against latitude, slope, SRI, and distance from moderate or high severity fire patches within eight previous years, measured from the center point of the transect. We also included forest structure covariates: percent tree cover of trees > 5 m tall, tree basal area, largest tree dbh, and maximum number of snags per hectare. We verified that none of these variables were correlated at higher than R > 0.55. All covariates were standardized prior to analysis, and we also included squared values of slope, tree cover, basal area, tree dbh, and snags to account for potential associations with intermediate or extreme values of these covariates. We selected a final model starting with a complete model including all 14 variables listed above via removal of the variable with lowest significance in a stepwise fashion until the AIC value (Akaike Information Criterion) ceased to improve (Burnham and Anderson 2002).
We initially included forest type (as a 6-level factor) in the static habitat association model but the model performed poorly and thus we eliminated this categorical variable from the stepwise model selection process. This was not surprising because topographical variables such as elevation, latitude, slope, and aspect are among the primary drivers of the distribution patterns of these broad forest types in the Sierra Nevada. Thus, we analyzed a separate static model using the same data structure with only the six broad forest types (hereafter referred to as the “static forest type model”).
The second goal of our occupancy analysis was to evaluate the proportion of area occupied by Black-backed Woodpecker over the three years in our study area. We used a multiseason dynamic model (hereafter referred to as the “dynamic model”), which includes probabilities of transect colonization and extinction between seasons, in our case, years. The time intervals between secondary survey events (within a single season) were up to six weeks, while the time interval between primary survey events was one year. We included the occupancy covariates identified from the lowest AIC value static habitat association model, because with our relatively low number of Black-backed Woodpecker detections we sought to avoid overparameterization of the model.
All models were analyzed using R version 3.0 statistical software and the package “unmarked” (Fiske and Chandler 2011, R Core Team 2013). All counts were converted to detection/nondetection (1 or 0). Both occupancy and probability of detection were defined by logit-linear models. Probability of detection in all models was evaluated as a function of an intercept term, and a covariate for survey type, passive  or playback . We defined the model for occupancy probability as the logit-transformed probability of occupancy in relation to the covariates listed above. In the dynamic occupancy model, colonization logit(γi) and extinction logit(εi) were assumed to be constants because we did not have enough observation data to accommodate more covariates.
We plotted the relationships between occupancy and covariates included in the final static habitat association model using the “predict” function in unmarked, by deriving predicted occupancy estimates from a range of values for the variable of interest and setting all other variables to mean values, except for elevation. We plotted the elevation residual at the third quartile value to produce a figure that shows the occupancy relationship at levels not truncated by extremely low predicted occupancy values. We plot all results as estimates with 95% confidence intervals (CIs).
We derived annual occupancy estimates from the dynamic model using the “smoothed” estimator in the R package “unmarked” and generated standard errors for occupancy estimates using 1000 nonparametric bootstraps (Weir et al. 2009, Kery and Chandler 2012). Turnover probability was calculated from the dynamic occupancy model colonization and extinction parameters and the derived occupancy estimates as described in Weir et al. (2009). We followed the approach of Kery and Chandler (2012), where the probability of a transect changing from occupied to unoccupied or vice versa between years is a function of the colonization probability in the previous year, the extinction probability in the previous year, and the proportion of unoccupied sites in the previous year. We report the two yearly turnover probabilities and estimated 95% CIs using standard error estimates from 1000 parametric bootstraps (Kery and Chandler 2012).
We detected Black-backed Woodpeckers on 75 of 386 (naïve transect detection rate = 19%) transects visited across the three years of our study (Fig. 1). We detected them on 27 of 373 (7%) transects surveyed in 2011, 38 of 373 (10%) surveyed in 2012, and 35 of 372 (9%) surveyed in 2013. They were detected in more than one year at 22 of the 75 (31%) green forest transects with detections.
The final static habitat association model contained seven covariates on occupancy: latitude, elevation, slope, SRI, maximum number of snags, snags², and largest tree dbh. The strongest associations were with physiographic variables: high elevations, northern latitudes, and low slopes (Table 1, Fig. 2). Very few detections were recorded below -200 m standardized elevation (approximately 1500 m elevation in the north and 1800 m in the south), and none south of 37.2° latitude. Slightly weaker associations were apparent with moderate to high snag densities, large diameter trees, and more northerly aspects (Figs. 2, 3).
The static forest type model showed that Black-backed Woodpecker occupancy was highest in the high elevation forest types, especially lodgepole pine (0.37), followed by red fir (0.23), white fir (0.20), and eastside-Jeffery pine (0.17; Fig. 4). Occupancy in lower elevation forest types including Sierra mixed conifer and other habitats was < 0.05.
Estimates of occupancy from the dynamic model was 0.21 in 2011 (95% CI: 0.11 0.31), in 2012 it was 0.20 (95% CI: 0.12 0.28), and in 2013 it was 0.19 (95% CI: 0.10 0.27). Colonization probability was 0.04, and extinction probability was 0.19 (Fig. 5). The probability of a transect changing from occupied to unoccupied or vice versa (turnover rate) from 2011 to 2012 was 0.13 (95% CI: 0.04 0.23) and from 2012 to 2013, it was 0.14 (95% CI: 0.05 0.24).
Transect level detection probability in the dynamic model for passive surveys was 0.20 (95% CI: 0.15 0.27) and 0.16 for the playback surveys (95% CI: 0.12 0.22). The combined detection probability for both survey types was 0.33 (95% CI: 0.25 0.43).
We found 19 nests in green forest between 2004 and 2013 at either the incubation or nestling stage. Nests were found across the entire study area, including east and west of the Sierra crest, the southern Cascades, the Modoc plateau and the southern Sierra. Two of the 19 nests occurred in the same tree (different cavity) in consecutive years. Nests ranged in elevation from 1524 to 2521 m. Five different tree species were utilized and 12 of the 19 nests were in dead trees. Lodgepole pine was the most frequently used (n = 5), followed by aspen (Populus tremuloides, n = 4), red fir (n = 3), Jeffrey pine (n = 3), whitebark pine (Pinus albicaulis, n = 1), and three unidentified tree species. Six of the seven live tree nests were in lodgepole or Jeffrey pine that had no visible evidence of decay. Seventy-nine percent of nests were found in lodgepole pine, red fir, or eastside pine habitat types. For the nine nests we recorded tree dbh, the average was 47 cm ± 1.8 SE (range 24 83 cm).
This study provides new information on the occupancy and habitat associations of a rare habitat specialist outside of its primary habitat across the Sierra Nevada region. Our findings suggest that Black-backed Woodpecker occurrence in green forest is far greater than has been suggested by others (Odion and Hanson 2013), though it is still much lower than what has been reported in burned forest in this region (Seigel et al. 2012), a pattern that is consistent with findings for the species elsewhere in its range (Goggans et al. 1989, Trebmlay et al. 2009, Rota 2013). Their distribution in green forest varies based on forest type, physiographic parameters, and habitat conditions, some of which follow patterns found in burned forest in the region. Conservation strategies for Black-backed Woodpecker, especially because they are rare and potentially a keystone species, should consider its green forest habitat in the Sierra Nevada and throughout its extensive range.
Black-backed Woodpecker affinity for lodgepole pine is consistent with descriptions of this species’ distribution in green forest in California in the early part of the 20th century (Grinnell and Miller 1944) and in the Cascades region of Oregon (Goggans et al. 1989). Lodgepole pine occurs in association with meadow edges from the mid to upper elevations and as a distinct habitat type above 1800 m in the north and 2400 m in the south (Mayer and Laudenslayer 1988). Overall lodgepole pine forest occupies a relatively small proportion of forested habitat in the Sierra Nevada.
Occupancy was relatively high in other high-elevation conifer types, including red and white fir, that occupy a larger proportion of forested habitat in the Sierra Nevada than lodgepole pine. Occupancy estimates in each of these forest types was higher than our overall estimate. This was due to very low occupancy in Sierra mixed conifer forest types, which dominate the study area. Saracco et al. (2011) found similar occupancy patterns in burned lodgepole pine and red fir forest in the Sierra Nevada, although our results suggest their habitat selection may be more restricted in green forest. For example, Saracco et al (2011) detected Black-backed Woodpeckers at a far higher percentage of burned Sierra mixed conifer and ponderosa pine point count stations compared to what we found in green forest. In fact, we had no detections in ponderosa pine forest types across the three years of our study.
When we examined habitat associations beyond forest type, we found variation in green forest occupancy patterns at the transect scale was best explained by physiographic variables rather than habitat structure. Similarly, in burned forests of the Sierra Nevada, among the strongest predictors of occupancy were positive associations with elevation and latitude (Saracco et al. 2011). We detected very few individuals below 1700 m in the north and 1900 m in the south or below 37.2° latitude. This relationship with latitude is likely the result of the southern portion of our study area representing the southern extent of their North American range as ample high elevation fir and lodgepole pine habitat exists in the southern Sierra (Mayer and Laudenslayer 1988).
The association between habitat structure and occupancy was not as strong as associations with elevation, latitude, and lodgepole pine forest, but we did find an association with areas that contained larger trees and patches of high snag density. In a preliminary analysis, we had included the average number of snags from all five sampling locations for each transect, instead of the maximum, and found no relationship with occupancy. Considering the patchy nature of snags and a reconsideration of the species’ natural history, we decided to include the maximum snag count from any of the five sampling locations in the static habitat association model instead of average snag count. Because maximum snag count was a significant predictor of occupancy it suggests that current guidelines that call for retention of an average of 9.9 snags per hectare (4 per acre; U.S. Forest Service 2004a) may not be appropriate for maintaining habitat for Black-backed Woodpeckers in green forest. In our models, there was a positive association with patches of high snag density, with transect scale high counts averaging 27 per hectare (11 per acre). Dense patches of snags, unevenly distributed on the landscape, likely those created through wood-boring beetle outbreaks, may be far more important than an even distribution of suitable nesting trees across a territory. Other researchers have found associations with late seral characteristics in green forest, including high snag densities (Goggans et al. 1989, Setterington et al. 2000, Tremblay et al. 2009, 2010). Further study of the composition and distribution of dead and dying trees within their home ranges and a more detailed evaluation of their habitat associations would help fine tune management recommendations for their green forest habitat.
One hypothesis explaining Black-backed Woodpecker occurrence in green forest is that the majority of individuals detected were not within established territories, but rather were actively dispersing or seeking out new suitable burned habitat. Movement through the green forest matrix would be expected because burned habitat is only suitable for the species for a limited time (Saab et al. 2007, Saracco et al. 2011), and usually in isolated patches across the landscape. Black-backed Woodpeckers are known for their strong association with an abundance of wood-boring beetles, especially following fire (Dixon and Saab 2000). Thus, we consider our 4% colonization, 19% extinction, and 14% total turnover estimates in green forest to be low for a species adapted to exploiting irruptive food resources. Siegel et al. (2012) estimated yearly point-scale rates of colonization and extinction in burned forest in the Sierra Nevada at 7% and 57%. Indeed we would expect point-scale turnover rates would be biased higher than for the transect scale (Hanski 1998). High colonization rates in recent (1-3 years) fires and high extinction rates in older (5-8 years) fires would be expected as individuals colonize new habitat and then abandon it as habitat quality declines rapidly 6-8 years postfire (Saab et al. 2007, Nappi and Drapeau 2009, Saracco et al. 2011).
Though it is important to consider the limitations outlined above, our findings suggest that Black-backed Woodpecker occupancy may be more stable over time in green forest than in burned forest, and at a minimum, along with our sample of nests, provides strong evidence that their presence in green forest is not characterized by dispersal of birds from one burn to another. We advise caution in interpreting habitat quality in burned and green forest solely based on presence of nests and occupancy values. In the Black Hills of South Dakota, population growth rates in recently burned areas were positive compared to negative growth rates in mountain pine beetle-infested green forest (Rota 2013). However, the author suggested beetle-infested forest may buffer population declines when burned forest is unavailable. Further study of productivity, home range size, finer scale habitat selection, and foraging ecology of green forest-nesting individuals is needed to guide conservation efforts for the species.
Black-backed Woodpeckers’ role as a keystone cavity creator in burned forest of the Sierra Nevada is distinguished by their ability to excavate cavities in recently fire-killed trees (Tarbill 2010, Seavy et al. 2012). In burned forest, a number of species showed preference for utilizing Black-backed Woodpecker nest cavities (Tarbill 2010). Even though Black-backed Woodpeckers are far less common than Hairy and White-headed Woodpeckers (Picoides albolarvatus) across most of the Sierra Nevada, their role as a keystone cavity creator may extend beyond burned forest and they may create cavities preferred by some species in green forest as well. Unique among the conifer dwelling Sierra Nevada woodpecker species, we found a substantial portion of green forest nests excavated in live trees. Similarly, in the Cascades, 66% of nests in beetle-infested lodgepole pine stands were in live trees (Goggans et al. 1989). This pattern of excavating cavities in live conifer trees may have several important ecological implications. First, they may create cavities where snag resources are limited. Second, live trees in which cavities are excavated may be of greater long-term value as a cavity resource because they tend to stand much longer than the already decaying snags where other woodpeckers prefer to excavate (Bagne et al. 2008, Edworthy et al. 2012). Finally, at each of the live tree cavities we found, the bark had been removed from the area immediately below the cavity and resin wells had been excavated causing sap to weep down the tree bole, a behavior that has been reported for the species elsewhere in its range (Dixon and Saab 2000). Woodpeckers in the southern Cascade Mountains, including Black-backed Woodpecker, have been shown to carry fungus and bacteria on their bills associated with wood decay (Farris et al. 2004). By targeting live trees, inflicting substantial wounds, and potentially introducing pathogens, Black-backed Woodpeckers may play an important role in creating snag resources in green forest.
Occupancy methods are robust to variation in detection probability related to survey type or effort and thus can correct for false absences and imbalances in survey efforts in both space and time (MacKenzie et al. 2006). Despite the robust methods used in the occupancy framework, the estimates produced are limited by field methodology and data quantity (MacKenzie and Royle 2005). Because we conducted this research in the context of a multispecies monitoring program, we were not able to allocate extra field effort toward increasing our detections of Black-backed Woodpeckers, such as conducting extra playback surveys (Saracco et al. 2011) or extending survey length (Ibarzabal and Desmeules 2006). In eastern Canada, detection probability in burned vs. unburned forest varied, with significantly longer time until detection in mature unburned forest (Ibarzabal and Desmeules 2006). Our occupancy methodology did not account specifically for habitat induced variation in detection probability, but uncertainty is captured in confidence intervals. We likely could have improved confidence in occupancy estimates, and colonization and extinction rates by conducting additional visits or longer playback surveys. The decision on how to allocate limited resources, including the trade-offs between more visits versus greater spatial coverage is one that researchers should carefully evaluate when designing a study (Bailey et al. 2007). We feel that our analysis provides meaningful estimates to assess the green forest distribution of Black-backed Woodpecker on National Forest in the Sierra Nevada and is an example of how large-scale avian monitoring programs, can answer questions beyond their primary scope (Geupel et al. 2011, Hutto 2005).
Threats to Black-backed Woodpecker populations in burned forest include salvage logging, fire suppression, and prefire reduction of forest fuels that reduces tree densities in future burned areas (Hutto 1995, Hoyt and Hannon 2002, Hutto 2008, Saab et al. 2009). Threats in green forests are rarely discussed and little studied, but may include salvage logging of beetle-killed stands (Bonnot et al. 2009), fuel-wood cutting, low snag retention rates (Tremblay et al. 2010), and potentially climate change.
Some conservation threats, including reduction of forest fuels and salvage logging, may be consistent across burned and green forest habitat, but climate change may have opposite effects. Climate change models predict large increases in burn severity and area burned in the Sierra Nevada (Miller and Urban 1999). These predicted changes should increase burned forest habitat for Black-backed Woodpeckers (Bond et al. 2012, Malleck et al. 2013). At the same time, climate models predict a significant reduction in high-elevation lodgepole pine and red fir forests by the end of the century (Lenihan et al. 2008). Especially in the northern Sierra Nevada, where maximum elevations are lower, the distribution of these forests may become reduced as they are pushed off the top of the mountain (Lenihan et al. 2008). In the southern Sierra Nevada, high elevation lodgepole pine and red fir forest may persist longer as the higher maximum elevations provide space for these habitats to move upslope. Although we found more Black-backed Woodpeckers in the northern Sierra Nevada national forests, the southern forests and national parks may harbor the majority of suitable green forest types preferred by this species by the latter half of this century (Stralberg et al. 2009). Today, the southern Sierra has considerable lodgepole pine and red fir forest yet Black-backed Woodpeckers have higher occupancy at northerly latitudes suggesting they may not readily expand their range south. With increasing effects of climate change, we may find the geographically isolated Sierra population stuck between a rock (declining habitat in north) and a hard place (southern extent of range).
We recommend secondary habitats should not be ignored in conservation planning for Black-backed Woodpeckers in the Sierra Nevada and likely elsewhere. In a broader sense, traditional approaches of focusing on primary habitat for rare species may need to be reassessed in the face of rapidly changing conditions that result in large shifts in the climate envelopes of species and their preferred habitats, changes in disturbance regimes, as well as other increasing threats.
We’d like to thank D. Craig and C. Howell of the USDA Forest Service for their support of this project. We wish to acknowledge our field crews for their tremendous dedication to data collection, particularly B. Campos, K. Etzel, and J. Tietz for coordinating field operations. We are also grateful for the assistance of forest staff throughout the Sierra Nevada for providing logistical support. N. Seavy provided numerous suggestions throughout the process that greatly improved the final product. Funding for this research was provided by the USDA Forest Service, Pacific Southwest Region. This is Point Blue Contribution number 1872.
Aldridge, C. L., and M. S. Boyce. 2007. Linking occurrence and fitness to persistence: habitat based approach for endangered Greater Sage-Grouse. Ecological Applications. 17:508-526. http://dx.doi.org/10.1890/05-1871
Bagne, K. E., K. L. Purcell, and J. T. Rotenberry. 2008. Prescribed fire, snag population dynamics, and avian nest site selection. Forest Ecology and Management 255:99-105. http://dx.doi.org/10.1016/j.foreco.2007.08.024
Bailey, L. L., J. E. Hines, J. D. Nichols, and D. I. MacKenzie. 2007. Sampling design trade-offs in occupancy studies with imperfect detection: examples and software. Ecological Applications 17:281-290. http://dx.doi.org/10.1890/1051-0761(2007)017[0281:SDTIOS]2.0.CO;2
Bond, M. L., R. B. Siegel, and D. L. Craig. 2012. A conservation strategy for the Black-backed Woodpecker (Picoides arcticus) in California. Version 1.0. The Institute for Bird Populations and California Partners in Flight, Point Reyes Station, California, USA. [online] URL: http://www.birdpop.org/DownloadDocuments/BBWO_Conservation_Strategy_for_CA_V1.0.pdf
Bonnot, T. W., J. J. Millspaugh, and M. A. Rumble. 2009. Multi-scale nest-site selection by Black-backed Woodpeckers in outbreaks of mountain pine beetles. Forest Ecology and Management 259:220-228. http://dx.doi.org/10.1016/j.foreco.2009.10.021
Burnham, K. P., and D. R. Anderson. 2002. Model selection and multimodel inference: a practical information-theoretic approach. Second edition. Springer, New York, New York, USA.
California Department of Fish and Game. 2012. Public notice: Black-backed Woodpecker candidacy. California Department of Fish and Game, Sacramento, California, USA.
Dixon, R. D., and V. A. Saab. 2000. Black-backed Woodpecker (Picoides arcticus). In A. Poole, editor. The birds of North America. Cornell Lab of Ornithology, Ithaca, New York, USA. http://bna.birds.cornell.edu/bna/species/509/articles/introduction
Edworthy, A. B., K. L. Wiebe, and K. Martin. 2012. Survival analysis of a critical resource for cavity-nesting communities: patterns of tree cavity longevity. Ecological. Applications 22:1733-1742. http://dx.doi.org/10.1890/11-1594.1
Farris, K. L, M. J. Huss, and S. Zack. 2004. The role of foraging woodpeckers in the decomposition of ponderosa pine snags. Condor 106:50-59. http://dx.doi.org/10.1650/7484
Federal Register. 2013. Endangered and threatened wildlife and plants; 90-day finding on a petition to list two populations of Black-backed Woodpecker as federally endangered or threatened. Docket No. FWS-R8-ES-2013-0034; 4500030114. Vol. 78, No. 68.
Fiske, I., and R. Chandler. 2011. unmarked: An R package for fitting hierarchical models of wildlife occurrence and abundance. Journal of Statistical Software 43:1-23.
Geupel, G. R., D. Humple, and L. J. Roberts. 2011. Monitoring decisions: not as simple as they seem? Trends in Ecology and Evolution 26:107 http://dx.doi.org/10.1016/j.tree.2010.11.007
Goble, D. D., J. A. Wiens, J. M. Scott, T. D. Male, and J. A. Hall. 2012. Conservation-reliant species. BioScience 62:869-873. http://dx.doi.org/10.1525/bio.2012.62.10.6
Goggans, R., R. D. Dixon, and L. C. Seminara. 1989. Habitat use by Three-toed and Black-backed Woodpeckers. General Technical Report 87-3-02. U.S. Forest Service, Deschutes National Forest, Bend, Oregon, USA.
Grinnell, J., and A. H. Miller. 1944. The distribution of the birds of California. Pacific Coast Avifauna No 27.
Hanski, I. 1998. Metapopulation dynamics. Nature 396:41-49. http://dx.doi.org/10.1038/23876
Hanson, C. T., K. Coulter, J. Augustine, and D. Short. 2012. Petition to list the Black-backed Woodpecker (Picoides arcticus) as threatened or endangered under the federal Endangered Species Act. Center for Biological Diversity, Tucson, Arizona, USA.
Hanson, C. T., and B. Cummings. 2010. Petition to the state of California Fish and Game Commission to list the Black-backed Woodpecker (Picoides arcticus) as Threatened or Endangered under the California Endangered Species Act. John Muir Project, Cedar Ridge, California, USA, and Center for Biological Diversity, Tucson, Arizona, USA.
Howe, R. W., G. J. Davis, and V. Mosca. 1991. The demographic significance of ‘sink’ populations. Biological Conservation 57:239-255. http://dx.doi.org/10.1016/0006-3207(91)90071-G
Hoyt, J. S., and S. J. Hannon. 2002. Habitat associations of Black-backed and Three-toed Woodpeckers in the boreal forest of Alberta. Canadian Journal of Forestry Research 32:1881-1888. http://dx.doi.org/10.1139/x02-109
Hutto, R. L. 1995. Composition of bird communities following stand-replacement fires in Northern Rocky Mountain (U.S.A) conifer forests. Conservation Biology 9:1041-1058. http://dx.doi.org/10.1046/j.1523-1739.1995.9051033.x-i1
Hutto, R. L. 2005. Northern region landbird monitoring program: a program designed to monitor more than long-term population trends. Pages 956-962 in C. J. Ralph, and T. D. Rich, editors. Bird conservation implementation and integration in the Americas: Proceedings of the Third International Partners in Flight Conference. 20-24 March 2002, Asilomar, California, USA. General Technical Report PSW-GTR-191. U.S. Forest Service, Pacific Southwest Research Station, Albany, California, USA.
Hutto, R. L. 2008. The ecological importance of severe wildfires: some like it hot. Ecological Applications 18:1827-1834. http://dx.doi.org/10.1890/08-0895.1
Ibarzabal, J., and P. Desmeules. 2006. Black-backed Woodpecker (Picoides arcticus) detectability in unburned and recently burned mature conifer forests in north-eastern North America. Annales Zoologici Fennici 43:228-234.
Keating, K. A., P. J. P. Gogan, J. M. Vore, and L. R. Irby. 2007. A simple solar radiation index for wildlife habitat studies. Journal of Wildlife Management 71:1344-1348. http://dx.doi.org/10.2193/2006-359
Kery, M., and R. Chandler. 2012. Dynamic occupancy models in unmarked (unmarked R package vignette). Swiss Ornithological Institute, Sempach, Switzerland and USGS Patuxent Wildlife Research Center, Laurel, Maryland, USA. [online] URL: http://cran.r-project.org/web/packages/unmarked/vignettes/colext.pdf
Lenihan, J. M., D. Bachelet, R. P. Neilson, and R. Drapek. 2008. Response of vegetation distribution, ecosystem productivity, and fire to climate change scenarios for California. Climate Change 87(Suppl 1):S215-S230. http://dx.doi.org/10.1007/s10584-007-9362-0
MacKenzie, D. I., J. D. Nichols, J. A. Royle, K. H. Pollock, L. L. Bailey, and J. Hines. 2006. Occupancy estimation and modeling: inferring patterns and dynamics of species occurrences. Elsevier Academic, San Francisco, California, USA.
MacKenzie, D. I., and J. A. Royle. 2005. Designing occupancy studies: general advice and allocating survey effort. Journal of Applied Ecology 42:1105-1114. http://dx.doi.org/10.1111/j.1365-2664.2005.01098.x
Mallek, C., H. Safford, J. Viers, and J. Miller. 2013. Modern departures in fire severity and area vary by forest type, Sierra Nevada and southern Cascades, California, USA. Ecosphere 4(12):153. http://dx.doi.org/10.1890/ES13-00217.1
Martin, K., and J. M. Eadie. 1999. Nest webs: a community-wide approach to the management and conservation of cavity-nesting forest birds. Forest Ecology and Management 115:243-257. http://dx.doi.org/10.1016/S0378-1127(98)00403-4
Mayer, K. E., and W. F. Laudenslayer. 1988. A guide to wildlife habitats of California. Department of Fish and Game, Sacramento, California, USA.
McClure, C. J. W., and G. E. Hill. 2012. Dynamic versus static occupancy: How stable are habitat associations through a breeding season? Ecosphere 3:art60. http://dx.doi.org/10.1890/ES12-00034.1
Miller, C., and D. L. Urban. 1999. Forest pattern, fire, and climatic change in the Sierra Nevada. Ecosystems 2:76-87. http://dx.doi.org/10.1007/s100219900060
Miller, J. D., E. E. Knapp, C. H. Key, C. N. Skinner, C. J. Isbell, R. M. Creasy, and J. W. Sherlock. 2009b. Calibration and validation of the relative differenced normalized burn ration (RdNBR) to three measures of fire severity in the Sierra Nevada and Klamath Mountains, California, USA. Remote Sensing of Environment 113:645-656. http://dx.doi.org/10.1016/j.rse.2008.11.009
Miller, J. D., H. D. Safford, M. Crimmins, and A. E. Thode. 2009a. Quantitative evidence for increasing forest fire severity in the Sierra Nevada and Southern Cascade Mountains, California and Nevada, USA. Ecosystems 12:16-32. http://dx.doi.org/10.1007/s10021-008-9201-9
Murphy, E. C., and W. A. Lehnhausen. 1998. Density and foraging ecology of woodpeckers following a stand-replacement fire. Journal of Wildlife Management 62:1359-1372. http://dx.doi.org/10.2307/3802002
Nappi, A., and P. Drapeau. 2009. Reproductive success of the Black-backed Woodpecker (Picoides arcticus) in burned boreal forests: Are burns source habitats? Biological Conservation 142:1381-1391. http://dx.doi.org/10.1016/j.biocon.2009.01.022
Nielsen, S. E., G. B. Stenhouse, and M. S. Boyce. 2006. A habitat-based framework for grizzly bear conservation in Alberta. Biological Conservation 130:217-229. http://dx.doi.org/10.1016/j.biocon.2005.12.016
Norris, A. R., and K. Martin. 2010. The perils of plasticity: dual resource pulses increase facilitation but destabilize populations of small-bodied cavity-nesters. Oikos 119:1126-1135. http://dx.doi.org/10.1111/j.1600-0706.2009.18122.x
Odion, D. C., and C. T. Hanson. 2013. Projecting impacts of fire management on a biodiversity indicator in the Sierra Nevada and Cascades, USA: the Black-backed Woodpecker. Open Forest Science Journal 6:14-23.
R Core Team. 2013. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. [online] URL: http://www.R-project.org/
Ralph, C. J., J. R. Sauer, and S. Droege. 1995. Monitoring bird populations by point counts. PSW-GTR-149. U.S. Forest Service, Pacific Southwest Research Station, Albany, California, USA.
Raphael, M. G., and M. White. 1984. Use of snags by cavity-nesting birds in the Sierra Nevada. Wildlife Monographs 86:1-66.
Roberts, L. J., R. D. Burnett, A. M. Fogg, and G. R. Geupel. 2011. PRBO MIS final study plan and sampling protocols for Mountain Quail, Hairy Woodpecker, Fox Sparrow and Yellow Warbler. Contribution #1714 PRBO Conservation Science, Petaluma, California, USA. [online] URL: http://data.prbo.org/apps/snamin/uploads/images/bioreg/PRBO%20MIS%20Final%20Study%20Plan%20January2011.pdf
Rota, C. T. 2013. Not all forests are disturbed equally: population dynamics and resource selection of Black-backed Woodpeckers in the Black Hills, South Dakota. Dissertation. University of Missouri-Columbia, Columbia, Missouri, USA.
Saab, V. A., J. Dudley, and W. L. Thompson. 2004. Factors influencing occupancy of nest cavities in recently burned forests. Condor 106:20-36. http://dx.doi.org/10.1650/7485
Saab, V. A., R. E. Russell, and J. G. Dudley. 2007. Nest densities of cavity-nesting birds in relation to postfire salvage logging and time since wildfire. Condor 109:97-108. http://dx.doi.org/10.1650/0010-5422(2007)109[97:NDOCBI]2.0.CO;2
Saab, V. A., R. E. Russell, and J. G. Dudley. 2009. Nest-site selection by cavity-nesting birds in relation to postfire salvage logging. Forest Ecology and Management 75:794-804.
Safford, H. D., and K. P. Van de Water. 2013. Using fire return interval departure (FRID) analysis to map spatial and temporal changes in fire frequency on National forest lands in California. Research paper PSW-RP-266. U.S. Forest Service, Pacific Southwest Research Station, Albany, California, USA.
Saracco, J. F., R. B. Siegel, and R. L. Wilkerson. 2011. Occupancy modeling of Black-backed Woodpeckers on burned Sierra Nevada forests. Ecosphere 2(3):31. http://dx.doi.org/10.1890/ES10-00132.1
Seavy, N. E., R. D. Burnett, and P. J. Taille. 2012. Black-backed Woodpecker nest-tree preference in burned forests of the Sierra Nevada, California. Wildlife Society Bulletin 36:722-728. http://dx.doi.org/10.1002/wsb.210
Setterington, M. A., I. D. Thompson, and W. A. Montevecchi. 2000. Woodpecker abundance and habitat use in mature balsam fir forests in Newfoundland. Journal of Wildlife Management 64:335-345. http://dx.doi.org/10.2307/3803231
Siegel, R. B., M. W. Tingley, and R. L. Wilkerson. 2012. Black-backed Woodpecker MIS surveys on Sierra Nevada national forests: 2011 annual report. Institute of Bird Populations, Point Reyes Station, California, USA. [online] URL: http://www.birdpop.org/DownloadDocuments/2011_BBWO_MIS_Monitoring.pdf
Stephens, S. L., R. E. Martin, and N. E. Clinton. 2007. Prehistoric fire area and emissions from California’s forests, woodlands, shrublands, and grasslands. Forest Ecology and Management 251:205-216. http://dx.doi.org/10.1016/j.foreco.2007.06.005
Stevens, D. L., and A. R. Olsen. 2004. Spatially balanced sampling of natural resources. Journal of the American Statistical Association 99:262-278. http://dx.doi.org/10.1198/016214504000000250
Stralberg, D., D. Jongsomjit, C. A. Howell, M. A. Snyder, J. D. Alexander, J. A. Wiens, and T. L. Root. 2009. Re-shuffling of species with climate disruption: a no-analog future for California birds? PLoS ONE 4(9):e6825. http://dx.doi.org/10.1371/journal.pone.0006825
Tarbill, G. 2010. Nest site selection and influence of woodpeckers on recovery in a burned forest of the Sierra Nevada. Thesis. California State University, Sacramento, California, USA.
Theobald, D. M., D. L. Stevens, D. White, N. S. Urquhart, A. R. Olsen, and J. B. Norman. 2007. Using GIS to generate spatially balanced random survey designs for natural resource applications. Environmental Management 40:134-146. http://dx.doi.org/10.1007/s00267-005-0199-x
Tremblay, J. A., J. Ibarzabal, C. Dussault, and J.-P. L. Savard. 2009. Habitat requirements of breeding Black-backed Woodpeckers (Picoides arcticus) in managed, unburned boreal forest. Avian Conservation and Ecology 4(1): 2 [online] URL: http://www.ace-eco.org/vol4/iss1/art2/
Tremblay, J. A., J. Ibarzabal, and J.-P. L. Savard. 2010. Foraging ecology of Black-backed Woodpeckers (Picoides arcticus) in unburned eastern boreal forest stands. Canadian Journal of Forestry Research 40:991-999. http://dx.doi.org/10.1139/X10-044
U.S. Forest Service. 2004a. Sierra Nevada forest plan amendment: final supplemental environmental impact statement and record of decision. U.S. Forest Service, Pacific Southwest Region, Vallejo, California, USA.
U.S. Forest Service. 2004b. Existing vegetation CALVEG. U.S. Forest Service, Pacific Southwest Region, Remote Sensing Lab, McClellan, California, USA.
U.S. Forest Service. 2007. Sierra Nevada forests management indicator species: amendment. U.S. Forest Service, Pacific Southwest Region, Vallejo, California, USA.
U.S. Forest Service. 2012. California: vegetation burn severity. Edition 12_2. U.S. Forest Service, Pacific Southwest Region, Remote Sensing Lab, McClellan, California, USA.
U.S. Forest Service. 2013. California: fire return interval departure (FRID_2012). U.S. Forest Service, Pacific Southwest Region, Remote Sensing Lab, McClellan, California, USA.
Virkkala, R. 2006. Why study woodpeckers? The significance of woodpeckers in forest ecosystems. Annales Zoologici Fennici 43:82-85.
Weir, L., I. J. Fiske, and J. A. Royle. 2009. Trends in anuran occupancy from northeastern states of the North American Amphibian Monitoring Program. Herpetological Conservation Biology 4:389-402.