Sexual Dimorphism in Morphology and Plumage of Endangered Yuma Ridgway's Rails: A Model for Documenting Sex

Many applications in wildlife management require knowledge of the sex of individual animals. For example, sex determination is important for effective monitoring, especially in rare species that persist in small habitat patches or have captive populations reared for reintroduction. Differentiating between sexes is also important for basic science applications, such as studies of behavioral ecology and life history evolution. The extent of sexual dimorphism varies among species (Székely et al. 2007), and researchers must often rely on behavioral cues (e.g., differences in vocalizations or reproductive behaviors), genetic analysis, or invasive procedures (e.g., laparoscopy) to determine sex in species lacking obvious sexual dimorphism (Fairbairn et al. 2007; Shealer and Cleary 2007; Friars and Diamond 2011). Although these techniques can be effective, they are often inefficient, expensive, or invasive; hence, their utility for field applications is limited (Overton et al. 2009; Eilers et al. 2012), especially for endangered animals.

Researchers also may use statistical methods to determine sex of captured animals based on subtle morphological differences between males and females. Discriminant function analysis is one such statistical method that has been widely used to distinguish males and females in monomorphic birds (Phillips and Furness 1997; Alarcos et al. 2007; Mischler et al. 2015). Discriminant function analysis provides a linear combination of measurements that best differentiates between males and females based on a sample of known-sex individuals (Grimm and Yarnold 1995; Phillips and Furness 1997). Application of the resulting canonical classification function can predict the sex of unknown individuals (Alarcos et al. 2007; Dechaume-Moncharmont et al. 2011). In addition to morphometric measurements, subtle differences in the color of plumage and bare parts (e.g., mandibles, tarsi) can help determine sex in otherwise monomorphic species (Farmer and Holmgren 2000; Valdez and Benitez-Vieyra 2016). For example, electronic devices can quantify subtle intersexual differences in the plumage color of birds (Eaton 2005; Valdez and Benitez-Vieyra 2016).

The Yuma Ridgway's rail Rallus obsoletus yumanensis is a rare species of high conservation priority with a limited range in the southwestern United States and northwestern Mexico (Stevens and Conway 2019; Conway and Eddleman 2000; Eddleman and Conway 2020; Harrity et al. 2020). Indeed, the Yuma Ridgway's rail has been listed as federally endangered since 1967 (U.S. Fish and Wildlife Service [USFWS] 2010) pursuant to the U.S. Endangered Species Act (ESA 1973, as amended). It can be difficult to identify the sex of Yuma Ridgway's rails in the field because of their subtle sexual size dimorphism and lack of obvious sexual dichromatism (Eddleman and Conway 2020). Sex identification is further complicated because both sexes incubate (and thus develop brood patches), and the dense vegetation Yuma Ridgway's rails inhabit combined with their cryptic nature hinders behavioral observations (Eddleman and Conway 2020; Harrity and Conway 2020a). An efficient, noninvasive, and inexpensive method to determine the sex of Yuma Ridgway's rails could improve our understanding of population demographics and thus help improve management actions for this endangered species. As such, our objectives were to 1) evaluate morphological and color differences between sexes, and 2) develop an accurate discriminant function to reliably identify the sex of Yuma Ridgway's rails across their geographic range in the United States and Mexico.

We captured Yuma Ridgway's rails in 11 regions of their geographic range, including 9 study regions in the United States (in California, Arizona, and Nevada) and 2 study regions in Mexico (in Sonora) (Figure 1; Table 1). Elevation relative to sea level ranged from −69 m at Sonny Bono Salton Sea National Wildlife Refuge in California to 684 m at Ash Meadows National Wildlife Refuge in Nevada. Wetlands varied across the 11 study regions, ranging from small managed wetland parcels with tightly regulated water levels (e.g., managed wetlands at Imperial National Wildlife Refuge near Yuma, Arizona) to expansive emergent wetlands fed with agricultural drainage water (e.g., Salton Sea, California) and mangrove wetlands in coastal estuaries (e.g., Navopatia, Sonora, Mexico). Southern cattail Typha domingensis was the dominant wetland plant at most of the nine study regions in the United States, but chairmaker's bulrush Schoenoplectus americanus, California bulrush Schoenoplectus californicus, common reed Phragmites australis, and salt cedar Tamarix ramosissima also were present in many wetlands. Red mangrove Rhizophora mangle, black mangrove Avicennia germinans, and white mangrove Laguncularia racemosa were dominant plants at the two study regions in Mexico.

We captured Yuma Ridgway's rails via 3 methods: noose carpets (Harrity and Conway 2020b), single-trammel mist nets, and modified drop-door traps (Conway et al. 1993; Bui et al. 2015). We measured seven morphometric traits for each bird (Figures S1 and S2, Supplemental Material; Table S1, Supplemental Material): natural wing chord (wing; ±0.5 mm), exposed culmen length (culmen; ±0.1 mm), tarsometatarsus length (tarsus; ±0.1 mm), middle toe length (toe; ±0.1 mm), tarsus plus middle toe (whole leg; ±0.1 mm), tail length (tail; ±0.5 mm), and body mass (mass; ±1 g). We collected three to five breast feathers for subsequent DNA analysis to confirm sex. We measured color in a three-dimensional color space, L*, a*, and b*, with a ChromaMeter-400 colorimeter (Konica Minolta Business Solutions U.S., Inc.). In this three-dimensional space, L* represents lightness (smaller values are darker), whereas a* and b* together indicate saturation and hue. High values of a* indicate more red and low values indicate more green. High values of b* indicate more yellow and low values indicate more blue. We collected nine color measurements on each bird, three replicates for each of three areas: the right upper breast (breast), base of lower mandible (mandible), and right greater coverts (coverts). We were only able to obtain color measurements for 91 of the 270 Yuma Ridgway's rails because either the plumage was too muddy or field and time logistics prohibited measuring color (Table S2, Supplemental Material). We chose the breast, mandible, and coverts for color measurements because they were conspicuous areas and tended to be brightly colored, suggesting they may inform individual quality during intersexual interactions (Andersson 1994; Olson and Owens 2005; Hill and McGraw 2006). Furthermore, these areas were mostly uniform in color (i.e., not barred like the rail's flanks) and had sufficient surface area to measure with the colorimeter. The same observer (EJH) took all color measurements to minimize observer-induced variation in measurements.

We extracted DNA from the breast feathers of 101 adult Yuma Ridgway's rails with a Qiagen DNeasy® Tissue Kit (Qiagen Inc., Valencia, CA). We amplified the sex-specific CHD genes (i.e., CHD-W and CHD-Z) with a polymerase chain reaction (PCR) following the methods outlined in Griffiths et al. (1998). We made the following adjustments to optimize the DNA amplification process for Yuma Ridgway's rails: we used P8 (5′-CTCCCAAGGATGAGRAAYTG-3′) and P2 (5′-TCTGCATCGCTAAATCCTTT-3′) primers and set thermocycling parameters to include an initial denaturation at 94°C for 15 min, followed by 35 cycles of denaturation at 94°C for 30 s, annealing at 48°C for 90 s, and extension at 72°C for 60 s, and then a final cool down at 4°C for 10 min. CHD-W is unique to female birds, whereas CHD-Z appears in both sexes. As such, females are heterozygous (ZW) and males are homozygous (ZZ). Because only two base pairs differentiated the CHD-W and CHD-Z genes in congeneric species (clapper rail Rallus crepitans and king rail Rallus elegans; Perkins et al. 2009), we added a fluorescent 6-FAM tag to the upstream primer and resolved the PCR products by electrophoresis on an ABI Prism Genetic Analyzer (Applied Biosystems, Inc., Foster City, CA). We verified the accuracy of this procedure for Yuma Ridgway's rails by first analyzing genetic samples from four individuals that we captured while gravid (and were thus known females). Moreover, we included one or more genetic samples from known-sex individuals in each batch of samples processed in the ABI Prism Genetic Analyzer as controls. We refer to the 101 Yuma Ridgway's rails with genetic results as genetically sexed individuals throughout this article. We consider the genetic results unambiguous but address possible errors in the discussion.

We analyzed color and morphometric data separately because we had only 16 Yuma Ridgway's rails that were genetically sexed and had a complete set of color and morphometric measurements. Hence, we first evaluated intersexual differences in morphology of genetically sexed individuals and used these birds to develop an optimal discriminant function to predict the sex of Yuma Ridgway's rails. Next, we evaluated intersexual differences in color and assessed the accuracy of discriminant functions fit with color measurements only. Finally, we applied the best discriminant functions to predict the sex of all Yuma Ridgway's rails that were not genetically sexed in our study.

We took two steps to produce a final color dataset for analysis. We first removed any single replicate of a color measurement that was ≥ 3 standard deviations away from the mean of all individual measurements because such extreme values probably represented misreadings. Misreadings were uncommon (0.6% of measurements) and did not exclude any individuals from further analyses. We then averaged the remaining measurements to create a mean color measurement for the three body locations on each Yuma Ridgway's rail. As such, each Yuma Ridgway's rail had up to nine color variables: mean L*_breast, mean a*_breast, mean b*_breast, mean L*_coverts, mean a*_coverts, mean b*_coverts, mean L*_mandible, mean a*_mandible, and mean b*_mandible. Some individuals had fewer than nine variables if they were too muddy or their plumage was too worn to collect color measurements. We evaluated differences in color measurements between males and females independently for each color variable with linear mixed models to account for unmodeled variation among study regions and potential seasonal effects on color (Paxton et al. 2010). That is, we fit linear mixed models with a color measurement as the response variable, sex and date of measurement (as Julian Day of year) as explanatory variables, and a random intercept for study region. We followed a similar protocol as with morphometric measurements to determine the most accurate discriminant function based on color measurements. However, sample sizes for color measurements varied because of missing measurements; thus, we limited model comparisons to ranking model accuracies against the most accurate morphometric-based models.

Discriminant function analysis produces an equation (i.e., discriminant function) to predict the sex of unknown individuals based on a single discriminant score. Individuals above a threshold discriminant score (i.e., cut-point) are classified as one sex, whereas individuals below the threshold are classified as the other sex. Discriminant function analyses estimate posterior probabilities of group assignment for each individual in the analysis, and we calculated the cut-point by first applying logistic regression with posterior probabilities as the response and discriminant scores as explanatory variables. We estimated the discriminant score where the posterior probability for both males and females was 0.5 (i.e., cut-point) from the fitted probability curve (Phillips and Furness 1997). Finally, we calculated the intercept for the final discriminant functions by summing the products of the scaling parameters (i.e., variable-specific coefficients analogous to regression coefficients) and the weighted-means of their associated measurements (Friars and Diamond 2011). We applied the resulting discriminant functions to predict the sex of all Yuma Ridgway's rails in our study, where we classified individuals with discriminant scores greater than the cut-point as females and individuals with discriminant scores smaller than the cut-point as males.

Geographic variation in morphology may limit the transferability of a discriminant function beyond the population for which it was derived (Shealer and Cleary 2007; Dechaume-Moncharmont et al. 2011). To test for geographic variation in Yuma Ridgway's rail morphology and sexual dimorphism, we regressed morphometric measurements against latitude (we included sex in these models to account for differences in the proportion of males and females at each study region). We used the MASS package (Venables and Ripley 2002) in R v4.0.2 (R Core Team 2020) for statistical analyses, and we used the ggplot2 package (Wickham 2016) to plot results. In addition, we used the ucpm function in the klaR package (Weihs et al. 2005) to assess discriminant function accuracies and the lme function in the nlme package (Pinheiro et al. 2021) to fit linear mixed models.

We captured 270 adult Yuma Ridgway's rails in Arizona, California, Nevada, and Mexico between 2016 and 2020 (Table S4, Supplemental Material). We genetically confirmed the sex of 51 males and 50 females and used their measurements to evaluate sexual size dimorphism. Males were significantly larger than females for all measurements (Figure 2; Table 2). Average sexual size dimorphism varied across measurements, with wing measurements showing the smallest difference between sexes (7.0%) and mass showing the greatest (18.1%), although mass showed the greatest within-sex variation (Table 2). Males had significantly redder mandibles (larger a* values) and darker greater coverts (smaller L* values) than females, but other color measurements were similar between males and females (Tables 3 and S5, Supplemental Material). Date affected mandible color more than plumage color. Indeed, all three color values for mandible showed significant effect of date, whereas covert brightness was the only plumage color measurement that showed a significant effect of date (Figure S3; Table 3).

Importantly, 13 of the 15 possible morphometric discriminant functions (different combinations of the morphometric measurements) classified genetically sexed Yuma Ridgway's rails with > 90% accuracy (Figure 3). Latitude (Lat) had a marginally significant effect on whole-leg length but had no effect on the other measurements (whole leg: β_Lat = 0.25, R = 0.76, P = 0.045; culmen: β_Lat = 0.16, R = 0.62, P = 0.06; wing: β_Lat = 0.12, R = 0.65, P = 0.44; tail: β_Lat = −0.13, R = 0.37, P = 0.23).

We developed discriminant functions from easily collected morphometric measurements that predicted the sex of Yuma Ridgway's rails with ≥ 95.6% accuracy, thereby providing investigators with reliable tools to determine the sex of Yuma Ridgway's rails from field measurements. Importantly, we demonstrate the accuracy of, and provide equations for, multiple discriminant functions to account for missing measurements and facilitate accurate sex determination after the breeding season when Yuma Ridgway's rails molt all remiges simultaneously (Eddleman and Conway 2020). Moreover, we quantified previously undescribed intersexual differences in the color of Yuma Ridgway's rails, thereby providing another potential means to determine sex in this rare bird.

Our discriminant functions performed comparably with those developed for congeneric species. Indeed, a discriminant function developed for California Ridgway's rails Rallus obsoletus obsoletus included tarsus, wing, and culmen measurements and predicted the sex of adult birds with 100% accuracy (Overton et al. 2009). Importantly, this discriminant function developed for California Ridgway's rails correctly predicted the sex of only 67.6% of the genetically sexed Yuma Ridgway's rails in this study. This low prediction accuracy is not surprising given that Yuma Ridgway's rails are smaller than California Ridgway's rails in all measurements (Eddleman and Conway 2020) and underscores the importance of developing discriminant functions for the species or population of interest. A discriminant function developed for clapper rails in Texas and Louisiana also included tarsus, wing, and culmen measurements and predicted the sex of adult birds with 100% accuracy (Perkins et al. 2009); we did not apply this function to predict sex of Yuma Ridgway's rails as the authors did not provide a complete formula. Finally, similar efforts to develop models for distinguishing sexes were less successful in Virginia rails Rallus limicola (Fournier et al. 2013). Our discriminant functions differed from the above-mentioned functions by including whole-leg measurements instead of tarsus measurements. Tarsus measurements were less repeatable (Table S3) and less accurate at predicting sex of Yuma Ridgway's rails than whole-leg measurements in our study and were thus excluded from our analyses.

Geographic variation in the morphology of a species may limit the transferability of a discriminant function beyond the population for which the function was developed (Shealer and Cleary 2007). Indeed, limited transferability (as demonstrated above) is a primary criticism of discriminant function analysis for sex determination (Dechaume-Moncharmont et al. 2011). We collected morphometric measurements from Yuma Ridgway's rails throughout their relatively small geographic range, which spans only three U.S. and two Mexican states. We found little evidence that Yuma Ridgway's rail morphology varied geographically, suggesting the discriminant functions we developed can be applied throughout their entire range. However, our sample sizes at the northern and southern extremes of the species range were small (n ≤ 13); thus, additional samples would help validate the ubiquity of our models and further evaluate potential geographic variation in this rare bird.

Species accounts describe Yuma Ridgway's rails as monomorphic, with no intersexual differences in color (Eddleman and Conway 2020); however, to our knowledge, this assumption had not been tested quantitatively before this study. Our results suggest that covert brightness and mandible redness differ significantly between sexes. Such differences may result from differences in diet or energetic demands during the breeding season (Figuerola and Senar 2005; Olson and Owens 2005; Price 2006), variable parasitic burdens (McGraw and Hill 2000), or sexual selection (Andersson 1994). Indeed, mandible color or covert brightness may signal individual quality during inter- and intrasexual interactions (Figuerola and Senar 2005; Senar 2006; Ferns and Hinsley 2008; Boves et al. 2014). We measured color at only three body locations (greater coverts, breast, and lower mandible), and our measurements did not cover ultraviolet wavelengths. As such, Yuma Ridgway's rails may also differ in color at locations on the body or in wavelengths that we did not measure. For example, juvenile brown-headed cowbirds Molothrus ater show sexual differences in the underwing coverts (Farmer and Holmgren 2000), and many apparently monomorphic avian species show strong sexual differences when evaluated at wavelengths across the full range of avian visual capabilities (Eaton 2005; Valdez and Benitez-Vieyra 2016). Variable light conditions during field measurements could have influenced our results. However, the Konica Minolta ChromaMeter-400 colorimeter is designed as a portable device for quantifying color of objects and has a “light-protecting tube” that excludes ambient light from the sensors when the device is pressed against the measurement surface. As such, we always pressed the device firmly against Yuma Ridgway's rails when collecting color measurements and collected three replicate measurements to reduce the influence of single outlier results. Moreover, variable light conditions could introduce noise into the data, but we have no reason to suspect they would bias results in any way (e.g., males and females were not measured under consistently different ambient light conditions). Finally, many of our color measurements varied with date; hence, more data are needed over a wider range of dates to fully explore the extent of sexual dichromatism in Yuma Ridgway's rails and the extent to which color measurements can be a useful method to determine sex in this species.

We used a common molecular sexing protocol (Griffiths et al. 1998) to identify the sex of Yuma Ridgway's rails. This protocol has been used successfully to identify sex of numerous rallid species (Shizuka and Lyon 2008; Fuertes et al. 2010; Eilers et al. 2012; Fournier et al. 2013), including the congeneric clapper rail and king rail (Perkins et al. 2009). However, molecular sexing techniques are not immune to errors. For example, American coots Fulica americana (Shizuka and Lyon 2008) and water rails Rallus aquaticus (Eilers et al. 2012) exhibit CHD-Z polymorphism that may lead to sexing errors (Dawson et al. 2001). As such, our genetic results may include < 100% accuracy; if so, the efficacy of the morphometric models may be slightly different from what we reported. However, we reduced the risk of genetic sexing errors by using an ABI Prism Genetic Analyzer to resolve PCR products and automatically interpret PCR results (Çakmak et al. 2017) and sexing only those Yuma Ridgway's rails that showed CHD-Z or CHD-W alleles matching the known sizes in this species (382 base pairs for CHD-Z and 380 base pairs for CHD-W). Moreover, we verified the accuracy of this technique by including genetic samples from known-sex Yuma Ridgway's rails (we documented complete agreement between genetic results and known sex). Finally, laboratory errors (e.g., contamination of PCR products) also may lead to spurious results. We followed strict laboratory protocols and included negative controls in all PCRs to minimize the risk of such issues. We believe our genetic results are accurate but need further research to evaluate the existence of CHD-Z polymorphism in this species.

Sex identification is often a critical component of ecological studies, particularly those that inform population monitoring and management efforts. We provide a broadly applicable discriminant function to improve sex identification in Yuma Ridgway's rails throughout their range. Our methods, whereby we combine molecular sexing techniques and field measurements to develop accurate discriminant functions for sex identification, can be readily adapted to other avian species of management concern. Furthermore, we report subtle, but significant, sexual dichromatism in Yuma Ridgway's rails. These results document sex-specific patterns in coloration that allow future researchers to test hypotheses to determine the mechanisms underlying sex-based differences in plumage coloration.

Please note: The Journal of Fish and Wildlife Management is not responsible for the content of functionality of any supplemental material. Queries should be directed to the corresponding author.

Table S1. Morphometric measurements for 270 adult Yuma Ridgway's rails Rallus obsoletus yumanensis captured across 11 study regions in Nevada, Arizona, California, and Sonora from 2016 to 2020.

Available: https://doi.org/10.3996/JFWM-20-095.S1 (30 KB CSV).

Table S2. Colorimeter measurements for 91 adult Yuma Ridgway's rails Rallus obsoletus yumanensis captured across 11 study regions in Nevada, Arizona, California, and Sonora from 2016 to 2020.

Available: https://doi.org/10.3996/JFWM-20-095.S2 (29 KB CSV).

Table S3. Repeatability of morphometric measurements estimated from 29 adult Yuma Ridgway's rails Rallus obsoletus yumanensis captured and measured more than once, either within the same year or across multiple years (2016–2020), at 11 study regions across their range in Nevada, Arizona, California, and Sonora. Included in the table for context is sexual size dimorphism, calculated as mean male measurements minus mean female measurements divided by male measurements. Indeed, repeatability varies across measurements, but the differences were all smaller than average sexual size dimorphism.

Available: https://doi.org/10.3996/JFWM-20-095.S3 (21 KB DOCX).

Table S4. Numbers of adult Yuma Ridgway's rails Rallus obsoletus yumanensis captured across all years (2016–2020) of the study at 11 study sites in Nevada, Arizona, California, and Sonora. We assigned sex via genetic methods or with the discriminant functions derived from the morphometric measurements of 101 genetically sexed Yuma Ridgway's rails (n = 51 males and 50 females).

Available: https://doi.org/10.3996/JFWM-20-095.S4 (22 KB DOCX).

Table S5. Mean color values for three body areas on adult Yuma Ridgway's rails Rallus obsoletus yumanensis. We collected data during 2016–2020 across 11 study regions in Nevada, Arizona, California, and Sonora. Values with an asterisk are significantly different between males and females.

Available: https://doi.org/10.3996/JFWM-20-095.S5 (22 KB DOCX).

Figure S1. We collected three measurements of the legs of Yuma Ridgway's rails Rallus obsoletus yumanensis captured in Nevada, Arizona, California, and Sonora from 2016 to 2020. We measured midtoe length (A) from the final scale at the proximal end of the middle toe (third digit) to the base of the toenail. We measured tarsometatarsus length (tarsus) from the notch at the lateral condyle of the tibiotarsus to the top of the foot (B). We measured whole-leg length from the back of the tibiotarsus to the base of the toenail of the middle toe (C).

Available: https://doi.org/10.3996/JFWM-20-095.S6 (2.19 MB TIF).

Figure S2. We measured the length of exposed culmen (A), wing-chord (B), and tail (C) from Yuma Ridgway's rails Rallus obsoletus yumanensis captured in Nevada, Arizona, California, and Sonora from 2016 to 2020.

Available: https://doi.org/10.3996/JFWM-20-095.S7 (2.4 MB TIF).

Figure S3. Date had a significant effect on four color measurements for Yuma Ridgway's rails Rallus obsoletus yumanensis captured in Nevada, Arizona, California, and Sonora from 2016 to 2020. Sex had a significant effect on the redness of mandibles (B) and greater covert brightness (C). Points represent measured values and lines represent predicted values. Date had no effect on the five other color measurements (not shown). n = 63 adult Yuma Ridgway's rails for (A), (B), and (C); n = 74 adult Yuma Ridgway's rails for (D).

Available: https://doi.org/10.3996/JFWM-20-095.S8 (2.99 MB TIFF).

Figure S4. Probability of being an adult male Yuma Ridgway's rail Rallus obsoletus yumanensis in relation to discriminant score based on the discriminant function with culmen, tail, and whole-leg measurements. We developed the discriminant function from measurements from Yuma Ridgway's rails with genetically determined sex (n = 51 males and 50 females). Dashed line represents the cut-point below which we classified Yuma Ridgway's rails as males. We misclassified one male as female with the model. Yuma Ridgway's rails were captured across their range in Nevada, Arizona, California, and Sonora from 2016 to 2020.

Available: https://doi.org/10.3996/JFWM-20-095.S9 (2.16 MB TIF).

Reference S1.[USFWS] U.S. Fish and Wildlife Service. 2010. Yuma clapper rail (Rallus longirostris yumanensis) recovery plan. Draft first revision. U.S. Fish and Wildlife Service, Southwest Region, Albuquerque, New Mexico.

Available: https://doi.org/10.3996/JFWM-20-095.S10 (1.24 MB PDF) and https://ecos.fws.gov/docs/recovery_plan/Draft%20Yuma%20Clapper%20Rail%20Recovery%20Plan,%20First%20Revision.pdf.

Funding for the genetic analyses was partially provided by Imperial National Wildlife Refuge (NWR). We received logistical support from the following organizations: Sonny Bono Salton Sea NWR; Imperial NWR; Cibola NWR; Havasu NWR; Bill Williams River NWR; Ash Meadows NWR; Salt River Project; Arizona Game and Fish Department; Nevada Department of Wildlife; California Department of Fish and Wildlife; and Pronatura Noroeste, AC. Numerous field technicians and volunteers assisted with fieldwork. Jennifer Adams and Michelle Keyes assisted with genetic analyses. W.R. Eddleman, two anonymous reviewers, and the Associate Editor provided valuable feedback on prior versions of this article. All work reported in this document was carried out in compliance with the following six permits: 1) Native Endangered and Threatened Wildlife Species Recovery Permit (TE-039466-3), 2) State of California Scientific Collecting Permit (SC-7187), 3) State of Arizona Scientific Collecting License (SP654041), 4) State of Nevada Scientific Collection Permit (39305), 5) University of Idaho Institutional Animal Care and Use Committee (2018-51), and 6) Federal Bird Banding Permit (22524).

Any use of trade, product, website, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government.