Desert bird associations with broad-scale boundary length: applications in avian conservation

Authors


Correspondence author. E-mail: kevin_gutzwiller@baylor.edu

Summary

  • 1Current understanding regarding the effects of boundaries on bird communities has originated largely from studies of forest–non-forest boundaries in mesic systems. To assess whether broad-scale boundary length can affect bird community structure in deserts, and to identify patterns and predictors of species’ associations useful in avian conservation, we studied relations between birds and boundary-length variables in Chihuahuan Desert landscapes. Operationally, a boundary was the border between two adjoining land covers, and broad-scale boundary length was the total length of such borders in a large area.
  • 2Within 2-km radius areas, we measured six boundary-length variables. We analysed bird–boundary relations for 26 species, tested for assemblage-level patterns in species’ associations with boundary-length variables, and assessed whether body size, dispersal ability and cowbird-host status were correlates of these associations.
  • 3The abundances or occurrences of a significant majority of species were associated with boundary-length variables, and similar numbers of species were related positively and negatively to boundary-length variables.
  • 4Disproportionately small numbers of species were correlated with total boundary length, land-cover boundary length and shrubland–grassland boundary length (variables responsible for large proportions of boundary length). Disproportionately large numbers of species were correlated with roadside boundary length and riparian vegetation–grassland boundary length (variables responsible for small proportions of boundary length). Roadside boundary length was associated (positively and negatively) with the most species.
  • 5Species’ associations with boundary-length variables were not correlated with body size, dispersal ability or cowbird-host status.
  • 6Synthesis and applications. For the species we studied, conservationists can use the regressions we report as working models to anticipate influences of boundary-length changes on bird abundance and occurrence, and to assess avifaunal composition for areas under consideration for protection. Boundary-length variables associated with a disproportionate or large number of species can be used as foci for landscape management. Assessing the underlying causes of bird–boundary relations may improve the prediction accuracy of associated models. We therefore advocate local- and broad-scale manipulative experiments involving the boundary types with which species were correlated, as indicated by the regressions.

Introduction

A landscape boundary is ‘a zone composed of the edges of adjacent ecosystems’ and an edge is ‘the portion of an ecosystem near its perimeter, where influences of the surroundings prevent development of interior environmental conditions’ (Forman 1995). For birds, boundaries between land-cover types have the potential to influence movements (Wiens 1995), serve as habitat (Vega Rivera et al. 1998), affect abundance of food (Burke & Nol 1998) and add complexity to the structure and floristic composition of vegetation (Yahner 1988). Roadside boundaries can influence birds by decreasing habitat quality and territory density (Ortega & Capen 1999), reducing interior habitat (Reed, Johnson-Barnard & Baker 1996) and adjoining a potential source of mortality (Forman et al. 2003). Land-cover and roadside boundaries can also attract nest predators and brood parasites (Gates & Gysel 1978; Paton 1994; Rich, Dobkin & Niles 1994). Through such effects, boundaries may influence the occurrence and abundance of bird species, which are important components of avian community structure.

Current understanding regarding the effects of boundaries on bird communities has originated largely from local- and broad-scale studies of forest–non-forest boundaries in mesic systems (Paton 1994; Flather & Sauer 1996; Woodward, Fink & Thompson 2001). Furthermore, we are not aware of any studies that have explicitly examined whether bird community structure is influenced by effects of both roadside boundaries and land-cover boundaries, yet these different boundary types occur together in almost all landscapes. Because stimuli from broad-scale environmental attributes are believed to trigger initial settlement or selection responses by birds searching for habitat (Svärdson 1949; Hildén 1965; Hutto 1985), assessment of the degree to which landscape-scale boundary conditions influence species’ occurrences and abundances is warranted. Studies in xeric environments containing both roadside and land-cover boundaries are needed to determine whether broad-scale boundary conditions affect landscape occupancy and hence the structure of desert bird communities.

Knowledge of broad-scale boundary variables that have disproportionate or large effects on the structure of an avian assemblage (a subset of species in a bird community) may be pivotal in identifying focal variables for landscape management and assemblage-level conservation. Efforts to identify biological traits that cause species to be sensitive to environmental conditions (Davies, Margules & Lawrence 2000; Henle et al. 2004) are valuable to conservationists because such traits may be useful as predictors of species’ occurrences and species’ responses to environmental change.

As a first step toward assessing whether bird community structure is related to broad-scale boundary length in deserts, and identifying patterns and predictors of assemblage-wide associations useful for landscape planning and management, we studied the relations between birds and six boundary-length variables in Chihuahuan Desert landscapes. Operationally, a boundary was the border between two adjoining land covers, and broad-scale boundary length was the total length of such borders in a large area. We tested for disproportionate and large effects of boundary-length variables on avian assemblage structure. To explore whether body size, dispersal ability and cowbird-host status are useful predictors of species’ associations with broad-scale boundary length, we assessed whether these traits were correlates of species’ relations with boundary-length variables.

Methods

study area

We collected data in Big Bend National Park (BBNP), a 3205-km2 area on the Texas–Mexico border (29°19′42″N, 103°12′21″W), during 1995–97. Lying within the subtropical Chihuahuan Desert, BBNP is covered by eight basic land-cover types: surface water (0·1%), developed areas (0·2%), bare ground (1·4%), montane woodland (2·6%), riparian vegetation (4·4%), limestone grassland (25·4%), igneous grassland (30·6%) and shrubland (35·3%). Details of the topography, altitudes, soils, rainfall, temperatures, definitions of land-cover types and characteristic plant species are available in Gutzwiller & Barrow (2001). BBNP is natural or semi-natural, with the latter conditions the result primarily of legacies of commercial ranching practised until 1944, when the Park was established. Human development in BBNP is sparse (Gutzwiller & Barrow 2003).

sampling-site establishment

We used roadsides as starting points for paths that led to the centres of sampling sites (Gutzwiller & Barrow 2003). Systematic sampling with a random start (Cochran 1977) was used to determine the locations of path starting points. Successive starting points were systematically placed so that centres of neighbouring sampling sites were ≥ 4 km apart. Because development density, road density and traffic volume in BBNP were quite low, average distances between sampling sites and roads (0·4 km) and between sampling sites and developed areas (10·8 km) were sufficient to prevent traffic noise from hampering aural-cue detection during bird sampling (Gutzwiller & Barrow 2003). We used the same set of sampling sites during 1995–97.

bird sampling

Each year, between late February and late May, a 12-week period that coincided with the breeding season in BBNP (Wauer 1996), observers conducted 20-min unlimited-distance point counts (Ralph, Sauer & Droege 1995), once each week, at the centres of 70 sampling sites. Data were collected under standard conditions of wind speed, air temperature and precipitation (Gutzwiller & Barrow 2003). Details of the techniques used to avoid potential time-of-day biases, enhance detection of flushed birds, avoid double counting of birds and preclude potential observer effects associated with clothing colour, bird-identification ability and the range of habitat types sampled can be found in Gutzwiller & Barrow (2003).

bird species’ traits

Factors that may affect broad-scale habitat selection influenced our consideration of body size, dispersal ability and cowbird-host status as potential predictors of species’ responses to broad-scale boundary length. As birds fly over landscapes, they may be able to sense broad-scale environmental features, including general boundary-length conditions. If broad-scale boundary length influences initial decisions in the habitat-selection process (Svärdson 1949; Hildén 1965; Hutto 1985), it has the potential to affect subsequent landscape occupancy by species and hence avian community structure. Through barrier or filter effects, boundaries may contribute to a landscape's resistance to animal movement (Forman 1995), which may constrain feeding, breeding and dispersal.

If smaller species cross boundaries less easily than larger species (Wiens, Crawford & Gosz 1985), smaller species may be more likely to avoid landscapes with high boundary length. If wide-ranging species perceive less heterogeneity and hence fewer boundaries than less-mobile species (Wiens 1992), stronger dispersers such as neotropical migrants may be less likely to avoid landscapes with high boundary length (cf. Bélisle & St Clair 2001). The brown-headed cowbird Molothrus ater is a brood parasite that specifically searches habitat edges for host nests, deposits its eggs in nests at or near boundaries, and causes host species to produce fewer young (Norman & Robertson 1975; Paton 1994; Askins 1995). Through impacts on host nest success, host species may be more likely to avoid landscapes with high boundary length than non-host species.

We used body mass (Dunning 1984) as an index of body size. Neotropical migrant status was used as an indicator of dispersal ability; a species was considered a neotropical migrant if any of its populations breed north of and winter south of the Tropic of Cancer (23°27′N) (DeGraaf & Rappole 1995). We used information in DeGeus & Best (1991) and in Ortega (1998) to determine whether a species was a host for the brown-headed cowbird. The bird species involved in the present analysis were native and typical of the habitat in BBNP.

boundary-length and landscape variables

For each sampling site, we used a global positioning system (GPS) with 0–1-m accuracy to obtain latitude and longitude coordinates. Land-cover data layers were derived from satellite images with 30-m resolution, and we used 1:24 000 scale TIGER digital line graph files for 1995 to develop data layers for roads; birds associate significantly with land-cover and road features measured at these resolutions in our study area (Gutzwiller & Barrow 2002, 2003). ArcView geographical information system software (ESRI 1998) was used to obtain landscape measurements.

Within a 2-km radius circle centred on each sampling site, we measured six boundary-length variables (Table 1). The 2-km radius circle is within the range of spatial extents for which significant bird–landscape relations have been detected in our study area (Gutzwiller & Barrow 2001) and in other regions (Van Dorp & Opdam 1987). All boundary lengths were measured to the nearest 0·1 km. Because road surfaces and verges were physically and biologically quite different from adjoining land-cover types, and the road surface separated the verges on either side of a road, we treated the two sides of a road as separate roadside boundaries. Reed, Johnson-Barnard & Baker (1996) argued that edge habitat exists on both sides of a road and applied this logic to estimate amounts of road-related edge habitat. Thus, lengths of roadside boundaries included both roadsides. When riparian vegetation bordered a river, as it did at several sites along the Rio Grande, lengths of riparian vegetation boundaries included riparian vegetation on both sides of the river.

Table 1.  Boundary-length and landscape variables measured within 2 km of sampling sites
VariableDefinition
TBLTotal boundary length (km), including boundaries between eight land-cover types (see Study area) plus roadside boundaries
LCBLength (km) of all land-cover boundaries between eight land-cover types (see Study area)
RDBRoadside boundary length (km)
SGBShrubland–grassland boundary length (km), igneous and limestone grassland combined
SRBShrubland–riparian vegetation boundary length (km)
RGBRiparian vegetation–grassland boundary length (km), igneous and limestone grassland combined
NLTNumber of types of land cover
DSPNumber of patches of shrubland
IGPNumber of patches of igneous grassland
LGPNumber of patches of limestone grassland
RVPNumber of patches of riparian vegetation
DSCShrubland coverage (%)
IGCIgneous-grassland coverage (%)
LGCLimestone-grassland coverage (%)
ELVAltitude at the site centre (m)

Boundary length may be influenced by other landscape variables, such as the number of land-cover types, the number of patches of different types and the percentage of the landscape occupied by different types. Because these conditions may also affect bird numbers (Freemark et al. 1995; Flather & Sauer 1996), control for such effects was necessary to determine whether boundary-length variables themselves were important in structuring bird assemblages. Road length, and therefore the length of roadside boundaries, may be associated with certain landscape features (e.g. altitude, number of land-cover types and number of patches of different types). Thus, to assess effects of boundary-length variables per se, landscape conditions associated with road length had to be accounted for as well.

To control for these various conditions, we included in our analyses landscape variables (Table 1) that co-varied with the six boundary-length variables (rS = 0·62–0·20, P≤ 0·0001–0·100, n= 70, a priori α= 0·10) or had the potential to affect bird assemblage structure in BBNP (Gutzwiller & Barrow 2001, 2002). We measured landscape coverage to the nearest 0·1%. We used a GPS to measure the altitude at the centre of the sampling site to the nearest 5 m. Through its association with temperature, rainfall and soil type, and thus with vegetation structure and floristic composition (Wauer 1971), altitude served as an integrated measure of a variety of environmental conditions. From maps (Barnes 1979) we derived digital geological data useful for identifying limestone and igneous grasslands.

statistical analyses

Mean abundance for a species was calculated as the mean number of individuals observed per count at a particular site during a given year. To study the probability of occurrence of a species, we analysed presence–absence data, which indicated whether or not a species was observed at least once during all of the counts for a particular site during a given year.

Biases in bird detection have the potential to influence assessments of bird–landscape relations (Gutzwiller & Barrow 2001). We controlled analytically for three conditions (hereafter extraneous variables) that were outside the focus of our study but had the potential to affect bird detection: observer identity (OBx, where x was an identifying number for an observer); whether or not an observer at the site centre could see farther than 100 m in all directions (BLK); and the number of weekly bird counts (NCS). We controlled statistically for variation in dependent variables associated with extraneous variables by including the latter in regression models when they were significantly correlated with dependent variables.

We used OBx to control analytically for effects that may arise from differences in observer experience or ability to detect birds. BLK was used to control for local conditions, such as tall or dense vegetation or abrupt topography near the site centre, that could block aural or visual detection of birds. At a few sites, observers were not able to sample during every week of the 12-week study season because of inclement weather or vehicle difficulties (Gutzwiller & Barrow 2003). For analyses involving probability of occurrence, we used NCS to control for variation in the detected presence of species associated with number of weekly counts. We did not use NCS as a variable in analyses of mean abundance because this variable's computation already included the number of weekly counts at a site.

The assemblage-level focus of our analysis compelled us to study as many bird species as possible. To detect some uncommon species, it was necessary to sample 12 times at each site using 20-min counts. Our need to include uncommon species, and our interest in drawing inferences for the entire breeding season, were not compatible with the assumptions and data requirements of other techniques used to control for species’ detection probabilities. Although these approaches are appropriate in many circumstances, the primary problems they presented for our situation were: untenable assumptions about detection of individuals and dynamics of site occupancy (capture–recapture, occupancy and removal-model methods); estimators that would have been impossible to apply or would have provided poor precision because of insufficient detections, particularly for uncommon species (distance and double-observer methods); and personnel demands that would have required us to use fewer sites, fewer sampling visits to sites or shorter count periods (double-sampling, double-observer and removal-model methods), all of which would have made it much more difficult to draw reliable inferences about uncommon species.

The details of these problems for our circumstances are presented in Gutzwiller & Barrow (2003). In short, we adjusted our dependent variables for species’ detection probabilities using an approach (statistical control for extraneous variables via regression) for which the assumptions were clearly met, and we were able to draw biologically meaningful inferences because our adjustment approach enabled us to include data from the entire breeding period for both common and uncommon species.

Spatial and temporal autocorrelations have the potential to inflate the statistical significance of regression coefficients. We computed spatial-trend variables, which were third-order polynomial terms based on the easting (E) and northing (N) for sites (Buckland & Elston 1993), and included these variables in regression analyses. These variables enabled us to control for broad-scale spatial trends, which is necessary before testing for site-to-site spatial autocorrelation (Kaluzny et al. 1998; Lichstein et al. 2002). We tested and adjusted for spatial autocorrelation using mixed-model procedures (Littell et al. 1996). Statistical problems caused by within-year autocorrelations were precluded by using a site (not a weekly bird count) as the unit of analysis, and problems caused by among-year autocorrelations were precluded by analysing data for each year separately. This latter step was also required to determine the number of years that species were associated with boundary-length variables (see below).

We used forward stepwise least-squares regression (SAS 1999) to determine whether a species’ mean abundance was related to explanatory variables (boundary-length variables, extraneous variables, landscape variables and spatial-trend variables). Species involved in analyses of mean abundance were those that were detected during a given year at ≥ 90% of the 70 sites. Uncommon species had so many zero abundances that the normality assumption for least-squares regression (Neter, Wasserman & Kutner 1989) could not be met. For these species we assessed relations between a species’ presence and absence (probability of occurrence) and explanatory variables with forward stepwise logistic regression (SAS 1999; Hosmer & Lemeshow 2000). For some logistic regression models, complete or quasi-complete separation occurred when species occupied < 20% or ≥ 90% of the 70 sites; this problem can generate invalid results (SAS 1999; Hosmer & Lemeshow 2000). Hence, logistic regression was used for a species only when it was present on ≥ 20% but < 90% of the 70 sites during a given year (Gutzwiller & Barrow 2001). We applied standard diagnostic and remedial methods (Neter, Wasserman & Kutner 1989; Hosmer & Lemeshow 2000) to preclude numerical problems (e.g. multicollinearity, separation, zero cell counts and imprecise regression estimates) and ensure that all assumptions of least-squares and logistic regression were met. We reported the relative importance of each explanatory variable in a model as the percentage of variation in the dependent variable that was associated with that variable; within a model, these percentages summed to the model R2.

Based on the regression results, we tallied the number of years during which a species’ mean abundance or probability of occurrence was positively or negatively associated with each boundary-length variable. We used one-tailed binomial tests (SAS 1999) to determine whether boundary-length variables were correlated with the abundances or occurrences of a significant majority (proportion > 0·50) of species in the assemblage, whether a significant majority of species was correlated positively, and whether a significant majority was correlated negatively (Hofor each test, proportion = 0·50; Zar 1999). Two-tailed binomial tests (Zar 1999) were used to assess whether each boundary-length variable's association with the assemblage was disproportionate to its length at our study sites. We used two-tailed binomial tests for paired-sample data (Zar 1999) to determine whether different boundary-length variables were related to significantly different proportions of species; paired-sample tests were necessary because the associations being compared in a given test involved the same species. The unit of analysis for binomial tests was a species, and all assumptions of these methods (Zar 1999) were met.

Kendall tau correlation coefficients (SAS 1999) were used to test whether the number of years during which species were associated with boundary-length variables was correlated with body size, dispersal ability and cowbird-host status. The positive or negative influence of a trait may vary depending on other traits and interactions among traits (Henle et al. 2004; Ewers & Didham 2006). In response to this possibility, and as a follow-up to our Kendall correlation analyses, we used Poisson and negative-binomial regression to explore whether actual relations were masked by co-variation or interactions among traits. We assessed relations between species’ associations with boundary-length variables and the main and interaction effects of the three traits; the number of years of positive relations and the number of years of negative relations with each boundary-length variable were analysed as separate dependent variables. The unit of analysis for these correlation and regression analyses was a species, and all assumptions of these methods (Allison 1999; Zar 1999) were met.

An a priori α= 0·10 was employed in all analyses to avoid type II errors (Gutzwiller & Barrow 2001). For a given least-squares or logistic regression model, we assessed the significance of each explanatory variable after using the sequential Bonferroni method (Rice 1989) to adjust a family-wide α= 0·10. This approach reduced the chance of inflated type I error rates (and overfitted regression models) that can result from multiple related tests (Miller 1981; Gutzwiller & Barrow 2001). This same adjustment was applied separately to determine the significance of the binomial tests for disproportionate effects, significance of the paired-sample binomial tests, significance of the Kendall correlation coefficients for a given biological trait, and significance of the explanatory variables in each Poisson and negative-binomial regression model.

For the least-squares and logistic regression analyses, we could not legitimately use confirmatory methods that presuppose a relatively small set of a priori candidate models (e.g. information-theoretic modelling; Burnham & Anderson 2002). Meaningful formulation of such models for each species and our suite of explanatory variables was not feasible because combined theoretical or empirical effects of the explanatory variables were not evident from the literature or other information, and influences of the individual variables on desert birds have not been studied sufficiently. Identification of candidate models from among the large number of potential models would therefore have been largely arbitrary. When relatively little is known about a system, the stepwise procedures we used are appropriate for initial analyses (Neter, Wasserman & Kutner 1989; Hosmer & Lemeshow 2000). Adequate theoretical and empirical information was also not available for variables involved in the binomial, Kendall correlation and Poisson and negative-binomial regression analyses. Thus, consistent with the first-step nature of this study, our analyses were exploratory.

Results

bird data

For each species, occurrence data are provided in Gutzwiller & Barrow (2001) and mean abundance data are presented in Gutzwiller & Barrow (2002). The assemblage we studied was composed of all landbird species that used ≥ 20% of the sampling sites each year. Among the 26 species involved, there was a wide range of abundance levels, reflecting inclusion of both common and uncommon species in our analyses. Percentages of species that were analysed using probability of occurrence and mean abundance were, respectively: 80·8%, 19·2% (1995); 76·9%, 23·1% (1996); and 65·4%, 34·6% (1997) (see Appendix S1 in the supplementary material).

boundary-length and landscape variables

The mean ± SE (range) for total boundary length, length of all land-cover boundaries, and roadside boundary length were, respectively: 132·7 ± 6·7 (10·7–261·6) km, 122·0 ± 6·7 (0·4–252·9) km and 10·6 ± 0·5 (2·8–29·9) km. Summary statistics for the other boundary-length and landscape variables are given in Gutzwiller & Barrow (2001, 2002). All of the variables had large ranges, indicating our analyses involved a wide variety of boundary-length and other landscape conditions.

bird–boundary relations and assemblage structure

For the 34 regression models containing significant boundary-length variables, R2 ranged from 12·8% to 79·6% (see Appendix S1 in the supplementary material). For this set of models, the mean amounts (and ranges) of variation in dependent variables associated with all boundary-length variables in a model were: 12·9% (5·1–28·3%) for 1995, 10·6% (3·4–28·5%) for 1996 and 8·2% (1·9–20·3%) for 1997 (see Appendix S2 in the supplementary material). Of the total amount of variation in dependent variables accounted for by each of the 34 models (represented by model R2), boundary variables accounted for the following mean percentages (and ranges) of variation: 37·4% (9·4–100·0%) for 1995, 27·0% (6·2–67·7%) for 1996 and 24·8% (4·4–71·2%) for 1997. Extraneous, landscape and spatial-trend variables were significant in many models, whereas site-to-site spatial autocorrelation was significant in only two models (1997 analyses for turkey vulture Cathartes aura and ladder-backed woodpecker Picoides scalaris). Model data for these two species for 1997 (see Appendices S1 and S2 in the supplementary material) reflect corrections for autocorrelation.

Considering both positive and negative effects together, a significant majority of species in the assemblage (proportion ± SE = 0·81 ± 0·08, P= 0·001) was associated with boundary-length variables during 1 or more years. However, when we considered positive and negative associations separately, a significant majority of species was not positively related (0·58 ± 0·10, P= 0·279) or negatively related (0·54 ± 0·10, P= 0·423) to boundary-length variables. Disproportionately small numbers of species were associated with total boundary length, length of all land-cover boundaries, and shrubland–grassland boundary length (variables responsible for large proportions of total boundary length), whereas disproportionately large numbers of species were associated with roadside boundary length and riparian vegetation–grassland boundary length (variables responsible for small proportions of total boundary length; Fig. 1).

Figure 1.

Each boundary-length variable's proportional contribution to total boundary length + SE, and each variable's proportional association with the bird assemblage (proportion of species whose abundances and occurrences were correlated with a boundary-length variable) + SE. Variable names are given in Table 1. For TBL, SE = 0 for the boundary-length proportion. P-values above bars are for two-tailed binomial tests for differences between a boundary variable's proportional contribution to total boundary length and its proportional association with the assemblage. All P-values, except that for SRB, were significant after a sequential Bonferroni adjustment of α= 0·10.

Roadside boundary length, related to six species positively and six species negatively (see Appendix S1 in the supplementary material), was associated with the largest proportion of the assemblage (0·46). Roadside boundary length was associated with a greater proportion of the assemblage than total boundary length, but none of the other comparisons of proportional associations with boundary-length variables was statistically significant (Table 2).

Table 2. P-values for two-tailed, paired-sample binomial tests for differences in the proportions of bird species associated with boundary-length variables. Column and row names for a given P-value indicate the two boundary-length variables involved in the comparison. See Table 1 for names of variables and Fig. 1 for proportions + SE. The bold P-value was significant after a sequential Bonferroni adjustment of α= 0·10
 TBLLCBRDBSGBSRB
LCB1·000    
RDB0·0040·013   
SGB0·1250·1090·344  
SRB1·0000·6880·0390·289 
RGB0·5080·2890·1800·7740·754

biological traits as predictors of bird–boundary relations

Body size, dispersal ability (neotropical migrant status), cowbird-host status and the number of years during which species were positively and negatively associated with each boundary-length variable are listed in Table 3. The number of years of negative associations with length of all land-cover boundaries was zero for all species, which prevented formal correlation tests involving this variable and species’ traits; nevertheless, the absence of negative associations with this variable implied no negative co-variation with the traits. Correlations between each trait and the number of years during which species were positively or negatively related to boundary-length variables were small in magnitude (Kendall's tau = -0·39–0·31) and none of the associated P-values (0·054–1·000) was significant after sequential Bonferroni adjustments. Results of the Poisson and negative-binomial regression analyses did not differ from the Kendall correlation results, indicating that co-variation or interactions among traits did not mask important relations.

Table 3.  Biological and statistical data used to assess whether body size, dispersal ability (neotropical migrant status) and cowbird-host status were correlates of species’ associations with six boundary-length variables (abbreviations in Table 1). The number of years during which a species’ abundance or occurrence was positively (+) or negatively (–) related to each boundary-length variable was derived from regression results (see Appendix S1 in the supplementary material). Brown-headed cowbirds do not build nests, so no data are provided for this species under Cowbird host. Names are from the American Ornithologists’ Union (1998). The US Fish and Wildlife Service (2002) and Partners in Flight (2005) indicate that species marked with an asterisk deserve regional, national or continental conservation attention
SpeciesBody mass (g)Neotropical migrantCowbird hostNumber of years with a significant relation
TBL (+, –)LCB (+, –)RDB (+, –)SGB (+, –)SRB (+, –)RGB (+, –)
Turkey vulture1467·0YesNo0, 00, 02, 00, 00, 00, 2
Cathartes aura
Scaled quail* 184·0NoNo0, 00, 00, 00, 00, 00, 1
Calipepla squamata
White-winged dove 153·0YesNo0, 00, 00, 00, 00, 01, 0
Zenaida asiatica
Mourning dove 119·0YesYes0, 00, 01, 00, 00, 00, 0
Zenaida macroura
Greater roadrunner* 376·0NoNo0, 00, 00, 10, 00, 02, 0
Geococcyx californianus
Lesser nighthawk  49·9YesNo0, 10, 01, 02, 00, 00, 0
Chordeiles acutipennis
Ladder-backed woodpecker  30·3NoNo0, 00, 00, 00, 00, 00, 0
Picoides scalaris
Say's phoebe*  21·2YesYes0, 01, 00, 00, 00, 00, 0
Sayornis saya
Ash-throated flycatcher  27·2YesNo0, 00, 00, 00, 10, 00, 0
Myiarchus cinerascens
Loggerhead shrike*  47·4YesYes0, 00, 00, 00, 00, 00, 0
Lanius ludovicianus
Bell's vireo*   8·5YesYes0, 00, 00, 00, 02, 00, 0
Vireo bellii
Common raven1199·0NoNo0, 01, 00, 00, 00, 00, 0
Corvus corax
Verdin*   6·8NoYes0, 00, 00, 00, 00, 00, 0
Auriparus flaviceps
Cactus wren*  38·9NoNo0, 00, 00, 00, 00, 00, 0
Campylorhynchus brunneicapillus
Rock wren  16·5YesYes0, 00, 00, 10, 00, 10, 0
Salpinctes obsoletus
Bewick's wren   9·9YesYes1, 00, 00, 10, 00, 00, 0
Thryomanes bewickii
Black-tailed gnatcatcher*   5·0NoYes0, 00, 00, 10, 00, 00, 0
Polioptila melanura
Northern mockingbird  48·5NoYes0, 00, 00, 11, 00, 00, 0
Mimus polyglottos
Canyon towhee*  44·4NoYes0, 10, 01, 01, 10, 00, 0
Pipilo fuscus
Rufous-crowned sparrow*  18·7YesYes0, 00, 01, 01, 00, 00, 0
Aimophila ruficeps
Black-throated sparrow*  13·5YesYes0, 00, 00, 01, 00, 00, 2
Amphispiza bilineata
Pyrrhuloxia*  35·5NoYes0, 00, 01, 01, 01, 00, 0
Cardinalis sinuatus
Blue grosbeak  28·4YesYes0, 00, 00, 00, 00, 00, 1
Guiraca caerulea
Brown-headed cowbird  43·9Yes0, 00, 00, 10, 00, 00, 0
Molothrus ater
Scott's oriole*  37·4YesYes0, 00, 00, 01, 00, 10, 0
Icterus parisorum
House finch  21·4NoYes0, 00, 00, 00, 00, 00, 0
Carpodacus mexicanus

Discussion

bird–boundary relations and assemblage structure

The abundance and occurrence aspects of assemblage structure were significantly related to boundary-length variables, substantiating that broad-scale boundary length can be an important determinant of avian assemblage structure in desert systems. These results emerged after we controlled analytically for various conditions (extraneous variables, landscape variables, spatial-trend variables and spatial autocorrelation) that had the potential to mask or spuriously enhance the significance of boundary-length variables. At the assemblage level, neither positive nor negative effects dominated associations with boundary-length variables, suggesting that the assemblage was not affected primarily by either positive phenomena (e.g. increased vegetation diversity, habitat and food; Yahner 1988; Vega Rivera et al. 1998) or negative phenomena (e.g. higher nest predation and nest parasitism; Gates & Gysel 1978; Paton 1994) that often occur along boundaries.

A disproportionately large number of species was associated with roadside boundary length, but the number of species that was positively and negatively associated with roadside boundary length was the same. Thus, a tendency for negative influences from roadside boundaries that might be expected from the literature (Rich, Dobkin & Niles 1994; Reed, Johnson-Barnard & Baker 1996; Ortega & Capen 1999) was not apparent in the desert assemblage we studied. Road surfaces and verges differed from adjoining land covers in terms of ground-surface conditions, vehicular disturbance, plant species composition and vegetation structure, suggesting that the absence of a tendency for negative influences was not because of similarities between roadsides and adjacent land covers. Species’ positive associations with roadside boundary length may have reflected use of roads or road verges as feeding sites (Poole 2005): turkey vultures feed on road-killed animals; mourning doves, canyon towhees and rufous-crowned sparrows feed primarily on seeds and may pick up roadside grit to aid digestion; lesser nighthawks catch insects on and above roads; and pyrrhuloxias feed on seeds along road verges. The causes of species’ negative associations with roadside boundary length may be more varied (Wauer 1996; Poole 2005): roads and associated traffic may have been too disturbing for greater roadrunners, Bewick's wrens and black-tailed gnatcatchers, which usually do not use highly developed areas; several hosts of the brown-headed cowbird (rock wren, Bewick's wren, black-tailed gnatcatcher and northern mockingbird) were negatively associated with roadside boundary length (see Appendix S1 in the supplementary material); in BBNP, northern mockingbirds often nest in floodplains, but most roads in BBNP lie outside the floodplains; and rock wrens typically nest on cliffs, in rock outcroppings and among fallen rocks, but in BBNP such areas typically have few or no roads.

Regarding the disproportionate association of the assemblage with riparian vegetation–grassland boundary length, the positive associations exhibited by white-winged doves and greater roadrunners may reflect their use of riparian areas for general habitat, nesting and a source of water (Wauer 1996; Poole 2005). Negative associations with this boundary type may have been because of the following factors (Wauer 1996; Poole 2005): scaled quail occupy shrublands where the plant density is probably lower than that along riparian vegetation–grassland boundaries in BBNP; turkey vultures may have been less abundant in larger riparian areas because such areas in BBNP are used by behaviourally dominant species (crested caracara Polyborus plancus, Harris’ hawk Parabuteo unicinctus and red-tailed hawk Buteo jamaicencis); black-throated sparrows occupy dry uplands and their eggs and nestlings are preyed on by greater roadrunners, whose occurrence was positively associated with riparian vegetation–grassland boundary length (see Appendix S1 in the supplementary material); and blue grosbeaks nest in larger riparian areas in BBNP, but they may have occurred less often in these areas because they typically nest low in vegetation, where their eggs and nestlings would be easy prey for several mammalian predators that use riparian areas in BBNP (Carter 2002).

biological traits as predictors of bird–boundary relations

We tested for positive and negative relations between species’ associations with boundary-length variables and each biological trait because knowledge about positive relations may be valuable for deciding how to manage landscapes to increase species’ abundance and occurrence, and information about negative relations may help to identify traits that cause species to be susceptible to boundary-length conditions. Although we did not find support for effects of body size, dispersal ability or cowbird-host status, analysis of these relations in other xeric systems is needed to determine whether our results apply to desert bird communities in general.

synthesis and applications

Boundaries are important components of landscapes because their effects on ecological processes are larger than one would expect based on their relatively small area in the landscape (Puth & Wilson 2001) and because they can influence bird movements (Wiens 1995), habitat quality (Ortega & Capen 1999) and quantity (Reed, Johnson-Barnard & Baker 1996), reproductive success (Paton 1994) and assemblage structure (see the Results). Furthermore, boundary length can be managed in some situations. For these reasons, information about effects of broad-scale boundary length may be valuable for avian conservation.

Many of the bird–boundary relations we observed are consistent with species’ habitat use as described in DeGraaf et al. (1991). For example, lesser nighthawks occupy dry areas with sparse vegetation, and during 2 years this species was positively associated with the length of shrubland–grassland boundaries. Bell's vireos associate with edges and require dense shrubs near water, and during 2 years this species was positively correlated with the length of shrubland–riparian vegetation boundaries. Pyrrhuloxias use dense vegetation along the edges of riparian areas, and during 1 year this species was positively related to the length of shrubland–riparian vegetation boundaries. Pyrrhuloxias also feed along field boundaries, and during 1 year this species was associated positively with the length of shrubland–grassland boundaries. These consistencies, and those described above concerning bird associations with roadside and riparian vegetation–grassland boundaries, indicate that local bird–habitat relations underlie the broad-scale patterns we observed. The consistencies also indicate that the regression models (see Appendix S1 in the supplementary material) are ecologically realistic starting points for understanding species’ relations with broad-scale boundary length and for applying these relations as working models (updated as new information becomes available) in avian conservation.

Of the 26 species studied, 13 have been identified as species that deserve conservation attention, and 10 of these 13 were significantly associated with one or more boundary-length variables (Table 3). Therefore, application of the bird–boundary relations has potential to provide conservation benefits. The regressions represent a first stage in the development of models that use boundary length to predict abundances and occurrences of individual species. Such models may be valuable for predicting consequences of boundary-length changes, and for predicting avifaunal composition in areas that conservation organizations are considering for acquisition.

Assessing the underlying causes of bird–boundary relations may help managers improve the prediction accuracy of associated models by clarifying the nature of predictors (linear or non-linear, additive or multiplicative) or identifying additional predictors. We therefore advocate experiments involving the boundary types with which species were correlated, as indicated by the regressions. Local-scale experiments at boundaries could manipulate food abundance, vegetation density (shelter, potential nest sites), brood-parasite numbers or nest-predator numbers; homing experiments with translocated territorial individuals (Bélisle & St Clair 2001) could be conducted to assess whether boundaries are barriers to bird movement. Because processes that occur at one scale may not be responsible for patterns at another scale (Wiens 1983), local-scale experiments should be complemented with broad-scale management experiments (Walters & Holling 1990). In protected desert areas such as BBNP, management efforts to restore grassland or riparian vegetation, to remove unnatural shrub communities or to remove roads could be designed as broad-scale experiments to assess causal effects of boundary-length variables on bird abundance and occurrence.

A focus on either land-cover boundaries or roadside boundaries alone would have prevented us from determining how the lengths of these boundary types were simultaneously associated with assemblage structure. To avoid unrealistic results and therefore potentially misdirected management efforts, relations with land-cover boundaries and roadside boundaries should be assessed simultaneously whenever possible.

Limited conservation resources often make it impractical to manage landscapes using a species-based approach. Ideally, management efforts should aid groups of species. Variables associated with a disproportionate or large number of species, such as roadside boundary length and riparian vegetation–grassland boundary length in BBNP, can be used as foci for landscape management.

Acknowledgements

We thank many colleagues (see Gutzwiller & Barrow 2003) who assisted with administrative, field and analytical activities. We are grateful to O. Schabenberger for statistical advice, S. H. Anderson and F. R. Gehlbach for suggesting manuscript clarifications, and J. F. Chace and B. D. Peer for cowbird information. This research was funded by the US Geological Survey (Biological Resources Division) via Cooperative Agreement 14-45-0009-94-1081.

Ancillary