Abstract
Brown bears Ursus arctos consume a wide range of organisms, including ungulates and plants, but Pacific salmon Oncorhynchus spp. are especially important to their diet where their ranges overlap. Although some brown bears minimize antagonistic encounters with other brown bears or infanticide by avoiding streams where salmon spawn, studies generally assume that brown bears with ready access to salmon feed heavily on them. To test this assumption, and the hypothesis that male brown bears would feed more heavily on salmon than females (owing to their sexual size dimorphism), we collected hair samples from brown bears by using barbed wire placed on six small tributaries of Lake Aleknagik, Alaska, USA, where adult Sockeye Salmon Oncorhynchus nerka are readily accessible and frequently consumed by brown bears. Analysis of DNA distinguished among the different brown bears leaving the hair samples, some of which were sampled multiple times within and among years. We assessed the contribution of salmon to the diet of individual brown bears by using carbon and nitrogen stable isotope signatures. The 77 samples analyzed from 31 different bears over 4 y showed isotopic ratios consistent with reliance on salmon, but the wide range of isotopic signatures included values suggesting variable, and in one case considerable, use of terrestrial resources. Stable isotope signatures did not differ between male and female brown bears, nor did they differ between two sides of the lake, despite marked differences in Sockeye Salmon density. We collected the hair samples when salmon were present, so there was some uncertainty regarding whether they reflected feeding during the current or previous season. Notwithstanding this caveat, the results are consistent with the hypothesis that salmon were sufficiently available to provide food for the brown bears and that the considerable isotopic variation among brown bears with access to salmon reflected their age, status, and behavior.
Introduction
In otherwise nutrient-poor terrestrial and aquatic ecosystems, the annual return of Pacific salmon Oncorhynchus spp. provides an important source of nutrients for a wide variety of organisms (Willson and Halupka 1995; Gende et al. 2002). Notwithstanding this diversity of consumers, brown bears Ursus arctos and black bears Ursus americanus are uniquely capable of catching and killing a large fraction of the salmon in small streams (Quinn et al. 2017), thereby influencing the distribution of marine nutrients in nearby terrestrial and aquatic ecosystems (e.g., Hilderbrand et al. 1999a; Meehan et al. 2005; reviewed by Helfield and Naiman 2006; Reimchen 2017; Harding et al. 2019). Salmon generally serve as a paramount prey source for bears because these anadromous fish are a rich source of energy, contributing to growth, reproductive success, and population density of bears (Farley and Robbins 1995; Hilderbrand et al. 1999b, 1999c; Fortin et al. 2007; Van Daele et al. 2013). However, bears are omnivorous with an extremely diverse diet (Mowat and Heard 2006; Seryodkin et al. 2016; Mangipane et al. 2018). For example, brown bears in the greater Yellowstone ecosystem consumed up to 266 different species, including plants, insects, birds, mammals, and fishes (Gunther et al. 2014).
The dietary diversity of brown bears reflects their need to consume nutrients for health, growth, and reproduction (Erlenbach et al. 2014) as well as the shifting spatial and temporal patterns of food availability (e.g., French et al. 1994). Prey availability and location can contribute to diet variability of brown bears (Hilderbrand et al. 1999b; Jones et al. 2006; Matsubayashi et al. 2014), and brown bears move to take advantage of the seasonal availability of specific foods such as alpine vegetation (Atwell et al. 1980), insects (French et al. 1994), Pacific Herring Clupea pallasii in intertidal areas (Fox et al. 2015), and Pacific salmon in streams (Deacy et al. 2018). Otherwise, brown bears may rely on prey such as ungulates (e.g., elk Cervus canadensis, caribou Rangifer tarandus, moose Alces; Mowat and Heard 2006) and plant material (Rode et al. 2001; Mowat and Heard 2006; Deacy et al. 2017; Mowat et al. 2017). Food availability may also vary from year to year, including salmon (Quinn et al. 2017) and berries (Deacy et al. 2017).
In addition to seasonal and annual variation in food resources, competition within and between bear species (Service et al. 2019) affects how much access individual bears have to spawning salmon (Gende and Quinn 2004). Male brown bears, being generally larger than the females, have greater access to meat (Hobson et al. 2000), including salmon (Rode et al. 2006; Van Daele et al. 2013; Adams et al. 2017). Female brown bears may also avoid feeding near streams with salmon to minimize the risk of infanticide (Ben-David et al. 2004), as part of a broader strategy to protect their young (Steyaert et al. 2013). However, brown bears of low social status can forage as successfully as more dominant brown bears by using alternative fishing tactics (Gill and Helfield 2012).
Several studies (e.g., Mowat and Heard 2006; Van Daele et al. 2013; Adams et al. 2017) made assessments of the variation in reliance on salmon by comparing stable isotope ratios of carbon (13C/12C) and nitrogen (15N/14N), from the hair of brown bears, to the isotopic signatures of salmon and alternative prey. In this study, we used stable isotope analyses to test the null hypothesis against two alternative hypotheses. The null hypothesis was that the isotopic signatures of brown bears would indicate reliance on salmon by all individuals because heavy predation on salmon (ca. 20–50% annually) occurs in streams within the home ranges of sampled brown bears (Quinn et al. 2017; Wirsing et al. 2018) and because partial consumption of salmon carcasses indicated that they are available in surplus (Lincoln and Quinn 2019; Lincoln et al. 2020a). The two alternative hypotheses suggest that samples might indicate 1) diverse diets, because brown bears tend to congregate on salmon streams from much wider ranges occupied earlier in the season (Glenn and Miller 1980; Sellers and Aumiller 1994), adding many other items to their integrated diets; and 2) less reliance on salmon and feeding on lower trophic levels by females than males, due to behavioral differences (Rode et al. 2006; Van Daele et al. 2013; Adams et al. 2017).
Information from our study system (detailed below) provided two alternative reasons to expect differences in foraging between male and female brown bears. First, we detected individual females more often than individual males (mean = 3.8 vs. 2.5 samples per animal; Wirsing et al. 2018). This finding might suggest that males were warier or otherwise behaved differently from females. However, videos indicated that females avoided the wires more often than did males (39 vs. 20%; Wold et al. 2020), suggesting greater wariness in females. Such differences in behavior might affect foraging patterns and access to salmon. Second, we sampled the difference in salmon abundance between the two groups of streams, and between years, thereby enabling us to assess an association between salmon abundance and higher dietary contribution from salmon.
Methods
We collected hair samples (n = 2,026) during July and August, 2012–2015, from six streams that flow into Lake Aleknagik (south central Alaska, USA, as part of the Wood River system): Happy, Hansen, and Eagle creeks on the northeastern side and Yako, Whitefish, and Bear creeks on the southwestern side (Quinn et al. 2014; Wirsing et al. 2018). We designate these two trios of streams as separate foraging neighborhoods because brown bears often foraged on two or three of the streams on one side of the lake, but rarely crossed the lake within or between years (Wirsing et al. 2018, and additional unpublished data). Annual estimates of Sockeye Salmon abundance in each stream, based on multiple counts each year (Quinn et al. 2017), indicated much higher densities on the northeastern side than the southwestern side (Table 1). To collect the hair samples, we strung two unbaited barbed wires (average, 8 m in length) with a barb every 12 cm across each stream and attached it to trees on either side by using fencing staples. We carefully checked the wires (ca. 50–55 cm above the streambed at midchannel in different reaches of each stream) for samples every second day to minimize sample degradation from prolonged exposure (Dumond et al. 2015). We removed hair samples from the wire with tweezers, placed the hair in dry coin envelopes in the field, and stored the sample with desiccant in the field. At the end of the season, we sent the samples to the Laboratory for Evolutionary, Ecological, and Conservation Genetics at the University of Idaho (Moscow, Idaho) for DNA analysis to verify the species (i.e., brown rather than black bear), identify the individual, and determine the sex of the bear.
We extracted genetic material from hair samples by using a DNeasy Blood and Tissue kit (Qiagen, Inc.). We generated brown bear genotypes for each sample by using 10 nuclear DNA microsatellite loci (Paetkau and Strobeck 1994; Paetkau et al. 1998; De Barba et al. 2010) and one sex identification marker (Ennis and Gallagher 1994). The observed and expected heterozygosities for these 10 loci are 0.71 and 0.69, respectively. We amplified each sample two to four times to ensure accuracy. We derived consensus genotypes following the rule that we must observe each allele twice at each locus and the genotype must contain data at eight or more loci to be included in the matching analysis. We considered two genotypes a match in the program GenAlEx (Peakall and Smouse 2006) if they were identical or included a one-allele mismatch at two or fewer loci that could be due to allelic dropout. Under this protocol, the probability of a match for unrelated individuals across 10 loci was 0.0000000011, and the probability of a match between siblings across 10 loci was 0.00025 (Waits et al. 2001). The probabilities of a match for unrelated individuals and siblings at eight loci ranged from 0.000000011 to 0.00000021 and from 0.00072 to 0.0017, respectively. For all single captures, we estimated the reliability of their genotype by using the program RelioType (Miller et al. 2002) and retained in the dataset if the reliability score was greater than or equal to 90%. For the years used in this analysis, the amplification success rate was 63%. We archived the genotype data in the Dryad Digital Repository (https://doi.org/10.5061/dryad.pnvx0k6kg).
We selected a total of 77 hair samples from 13 female brown bears and 18 male brown bears (mean = 2.5 samples per bear; range = 1–9), collected between 2012 and 2015, for stable isotope analysis (Table 2). The samples spanned the peak period of salmon spawning and brown bear activity at these sites: 18 July–27 August (Quinn et al. 2014; Wold et al. 2020). To the extent possible, we tried to represent both male and female brown bears from each of the six streams in different years. We rinsed the hair samples with 2:1 chloroform : methanol and distilled water to remove any debris and surface oils; we then oven dried the samples at 50°C for approximately 12 h. Next, the samples were cut to less than 1 cm in length, and we packed approximately 0.40 mg of hair into tin capsules for stable isotope analysis. We processed the samples at the University of Washington's IsoLab (Seattle, Washington) by using a Costech elemental analyzer, Conflo III interface, MAT253 isotope ratio mass spectrometer for a continuous flow-based measure for carbon (13C/12C) and nitrogen (15N/14N) stable isotope ratios. We report the data using standard delta notation that describes the per mil (‰) deviation in the ratio of heavy-to-light isotope relative to accepted international standards, in this case Pee Dee Belemite for carbon and air (N2) for nitrogen. The study accuracy was ±0.29 δ13C and ±0.61‰ for δ15N, based on standard deviations for internal laboratory standards (Salmon, GA1, GA2). Sample precision was ±0.27 for δ13C and ±0.21‰ for δ15N, calculated using sample duplicates. We differentiated visually between guard hair and underfur by using texture and length and primarily analyzed guard hairs (n = 74) because they integrate the annual diet (Jones et al. 2006). However, we detected no differences in stable isotope signature between three pairs of underfur δ13C = −20.87, δ15N = 9.20 and guard hair δ13C = −20.71, δ15N = 9.37 (P ≥ 0.30 in both cases), so we combined the samples for analysis.
We first used a linear mixed effects model by using the lme4 package in R Studio (Bates et al. 2015) to assess whether sex, foraging neighborhood, and collection day-of-year (starting on 1 July) influenced δ13C and δ15N signatures, compared with a null model (intercept only, without the predictor variables; Table S2, Supplemental Material). Bear identification (ID) was included as a random effect to account for individual variation and to allow for all values to be incorporated from bears sampled more than once. The linear mixed effects model took the form (R formula notation): δ13C or δ15N ∼ sex + foraging neighborhood + collection day-of-year + (1 | bear ID). We then used the Akaike Information Criterion (AICc) to identify top models (with ΔAICc < 2) from a candidate set for carbon and nitrogen isotope signatures of each individual bear (Akaike 1974).
We used isotopic data on Sockeye Salmon and locally available berries as dietary reference points with which to compare individual bear signatures (see also Hatch et al. [2019] for a similar approach for black bear trophic position). We did not attempt to survey all the possible prey items (e.g., both moose and caribou are found in the area) and build a complete mixing model. However, for reference purposes, stable isotope values for Sockeye Salmon from one of the six streams that we sampled (Hansen Creek) were reported by Johnson and Schindler (2013; also see Table 1), and we averaged those values to represent Sockeye Salmon available to these bears. In late July 2019, we collected samples of three species of berries along the streams sampled: cloud berry, also known locally as low-bush salmon berry Rubus chamaemorus; crow berry, locally known as blackberry Empetrum nigrum; and blueberry Vaccinium sp.). These berries are common near the streams and also often seen in bear scat (T.P. Quinn and A.J. Wirsing, personal observations). We corrected neither the salmon (means: δ13C = −20.59, δ15N = 11.26) nor the berry (blueberry: δ13C = −26.09, δ15N = −6.16; cloud berry: δ13C = −26.71, δ15N = −0.20; crow berry: δ13C = −25.78, δ15N = −7.87) samples for trophic enrichment (Stricker et al. 2015), and they are intended as raw comparison points.
Results
There was a strong positive correlation between δ13C and δ15N values, both when considering all samples (r = 0.93, P < 0.001) and the mean values for each (or single value, if only one was obtained; r = 0.91; Figure 1). One female brown bear (BEAR090), represented by three samples in the same year, showed much lower values of both isotopes (means: δ13C = −23.11, δ15N = 3.60) than any other brown bear (Figure 1; Table S1, Supplemental Material). The δ13C values ranged from −22.44 to −18.11 (mean = −23.11), and the δ15N values of the other individuals ranged from 8.85 to 14.60 (mean = 11.77).
Based upon an AICc criterion, the data most strongly support models for both δ13C and δ15N with only the collection day-of-year as a predictor and the random effect of brown bear ID (Tables S2 and S3, Supplemental Material). Collection day-of year estimates for δ13C and δ15N were −0.04 and −0.06, respectively. That is, the values tended to be lower later in the season than they were earlier in the season. The random effect variation for individuals was larger than that of the residual variation for both δ13C and δ15N. The other candidate models for both carbon and nitrogen with support (ΔAICc < 2) included sex and collection day-of-year as predictors (ΔAICc: δ13C = 1.863, δ15N = 1.245). One additional candidate model for nitrogen with support (ΔAICc < 2) included foraging neighborhood and collection day-of-year as predictors (ΔAICc: δ15N = 1.916).
Examination of the data from each sample confirmed the broad overlap in values between the sexes (δ13C: male average = −19.872 [SD = 0.934] vs. female average = −20.153 [SD = 1.380]; δ15 N: male average = 11.985 [SD = 1.204] vs. female average = 11.277 [SD = 2.668]). Similarly, there was broad overlap in values in creeks (mean δ13C: Happy, Hansen, Eagle = −19.927 [SD = 1.238] vs. −20.208 in Whitefish, Bear, Yako (SD = 1.210); mean δ15N: Happy, Hansen, Eagle = 11.670 [SD = 2.586] vs. Whitefish, Bear, Yako = 11.389 [SD = 1.636]). Year was not included in the formal model structure because it was uncertain whether the hair collected in midsummer reflects foraging that year or the previous year. However, higher δ13C and δ15N values (consistent with greater dietary reliance on salmon) were not detected in years when the salmon were more numerous, or the year after salmon were very abundant (Table 1).
We noted many brown bears represented by only one or two samples, but for some individuals we had multiple samples from the same or different years. Values for samples within the same year were characteristically very similar, but in some cases there was marked variation between years. One female (BEAR012; Figure 2A) sampled five times over 3 y showed intermediate values in 2012 (δ13C = −20.52, δ15N = 10.70), higher values in 2013 (δ13C = −19.30, δ15N = 12.70), and then lower values in 2015 (δ13C = −21.88, δ15N = 8.91). Other samples from individuals varied within a single collection year. For example, we analyzed seven samples from a female (BEAR005; Figure 2B) in 2013 with wide isotopic ranges (δ13C from −20.48 to −18.27, δ15N from 10.68 to 14.88). We collected the endpoints of this range within 4 d of each other in early August. Two samples that we analyzed from the same brown bear in 2014 were depleted outside of the range observed in the previous year (δ13C = −21.76 and −22.08, δ15N = 8.23 and 8.84).
Discussion
Although we obtained samples from brown bears on streams where many salmon are killed and eaten, in whole or in part, every year by bears (Quinn et al. 2017; Lincoln and Quinn 2019), we observed a wide range of isotopic signatures (Figure 2) in both carbon (δ13C from −23.11 to −18.37) and nitrogen (δ15N from 3.60 to 14.60). These ranges indicated that although some brown bears specialized, most brown bears consumed a mix of salmon and dietary items from lower trophic levels, consistent with the plant material commonly seen in bear scats along the streams (T.P. Quinn and A.J. Wirsing, personal observations). The correlation of the carbon and nitrogen values is consistent with brown bear diets in general (Hobson et al. 2000; Mowat and Heard 2006; Matsubayachi et al. 2014), although in some cases the relationship is not strong (e.g., Hobson et al. 2000; Hatch et al. 2019). Some bears had intermediate isotopic values between salmon and terrestrial items, and we infer that they obtained nutrients from both sources. However, other brown bears had apparently been more specialized, consuming either more or less salmon than the average. The coupling of nutrients by predators using different energy sources can contribute to general ecosystem productivity (Rooney et al. 2006).
We found that some of the hair samples were more enriched than the salmon on which the brown bears feed, likely because we did not adjust our sample values with trophic enrichment factors (defined by Stricker et al. [2015) as “the net isotopic difference between consumer tissues and prey, which is the summation of the various vital isotope effects related to digestion, assimilation, and routing”). These important factors are difficult to determine outside of very controlled conditions such as captive feeding studies. Lipid content and the isotopic value of the diet can strongly affect isotopic discrimination between brown bears and their food sources, and the output of mixing models may not represent the ecological patterns (Caut et al. 2009; Rode et al. 2016). We therefore chose to report unadjusted values because salmon and other potential food items for brown bears might differ greatly in lipid density and thus in discrimination factors.
The large variation in diet indicated in our data is not unusual for brown bears with access to salmon, but most previous studies encompassed very large (e.g., Adams et al. 2017; Mowat et al. 2017) or intermediate (e.g., Van Daele et al. 2012) spatial scales, whereas the separation of our focal streams in each neighborhood was just a few kilometers. The streams that we sampled are among the earliest creeks used by salmon in this system, although Whitefish and Eagle creeks are later than the other creeks (Lisi et al. 2013), and both hair collection and motion-activated camera images indicate that bears congregate on them as salmon arrive (Quinn et al 2014; Quinn and Wising, unpublished data). It is therefore unlikely that the brown bears had fed on salmon elsewhere earlier in the season. Rather, they probably fed on terrestrial food sources and then moved to the streams as salmon arrived in mid-to-late July, because the high densities of brown bears detected on the streams (Wirsing et al. 2018) would not be sustainable without salmon. This shift in diet is common in brown bears. For example, at the McNeil River State Game Sanctuary in Alaska, brown bears shown a dispersed density, feeding on sedges and other plants before the arrival of salmon, and then they congregate in large numbers at sites where salmon are easily caught (Sellers and Aumiller 1994).
We did not detect a difference in isotopic signatures between the groups of streams on the two sides of the lake, despite having previously documented markedly and consistently higher salmon densities on the side with Happy, Hansen, and Eagle creeks (Quinn et al. 2017; Table 2). Brown bears very seldom moved from one side of the lake to the other, even between years (Wirsing et al. 2018); thus, movement between neighborhoods cannot explain the similarity in isotope signatures. Rather, this pattern (and the failure to associate salmon abundance and isotope signatures among years) further suggests that in these streams, salmon are sufficiently available as prey for brown bears. All the streams are clear, shallow, and narrow, and long-term data reveal predation by brown bears averaging approximately 20–50% of the Sockeye Salmon present each year (Quinn et al. 2017). This interpretation, that salmon are sufficiently abundant to meet the needs of the brown bears, is also consistent with the prevalence of partial consumption of salmon by bears in this system (Lincoln and Quinn 2019; Lincoln et al. 2020a).
We also hypothesized that males would have isotopic values indicating more dietary reliance on salmon than females due to social dominance (e.g., Van Daele et al. 2013; Adams et al. 2017), but we detected no difference. This pattern is consistent with results of research elsewhere in Alaska (Fortin et al. 2007) and with the idea that salmon are sufficiently available and accessible in these streams to both sexes of brown bears, as evidenced by partial consumption of salmon carcasses (Lincoln and Quinn 2018; Lincoln et al. 2020a). Whatever enhanced access males might have through social interactions may be mitigated by some greater wariness compared with females. In these streams, we detected individual males less often than females (Wirsing et al. 2018). Moreover, the broad distribution of spawning along several kilometers of stream in each site likely prevents dominant bears from monopolizing access to the salmon. This broad distribution contrasts with well-studied sites such as the McNeil River falls, where bears congregate in close proximity and display alternative fishing tactics (Gill and Helfield 2012). Deacy et al. (2018) drew the same conclusion as did Lincoln et al. (2020a): beyond some threshold, increased salmon density may not substantially affect their consumption by brown bears, and variation in diet reflects individuality in brown bear behavior more than salmon availability.
There are two important methodological caveats to consider in interpreting our results. First, we could determine the sex, but not the age, of brown bears; so, the values likely included cubs, subadults, and adults, because all stages are detected by cameras paired with the wires (Lincoln et al. 2020b; Wold et al. 2020). Second, the brown bears were likely molting and growing new hair when we collected the samples; so, the data might reflect some mix of feeding from the current and previous summers (Mowat et al. 2017). Bears that we sampled repeatedly displayed either little intra-annual variation (Figure 2A) or intra-annual variation large enough that it is unlikely to be attributed to diet (Figure 2B). Our results suggest that in some years, both new- and old-growth hair was collected; this finding may explain why in some cases samples from the same year had very different isotopic signatures (Figure 2B; e.g., BEAR068). We probably collected both new- and old-growth hair from this brown bear in 2014, and old-growth hair was collected in 2015. Although collection timing accounts for variation in the δ13C and δ15N top models, we caution making conclusions on seasonal dietary patterns because of the possible extent of variation within a year for brown bears sampled repeatedly (Figure 2B). Ideally, all samples would have been collected in the fall, several months after the bears had fed on salmon, or in spring, before the arrival of salmon. However, at those times of the year, the bears would likely have been dispersed, and barbed wires on the streams would have been much less effective. Camera traps paired with the barbed wires on these six streams showed a pronounced increase in detections around the time when salmon arrived (mid-to-late July) and a decline at the end of August when salmon in these streams have largely completed spawning (Quinn et al. 2014), consistent with the tendency of brown bears to congregate around streams when spawning salmon are available (Berns et al. 1980; Glenn and Miller 1980). The alternative, baited hair snare stations (e.g., Dumond et al. 2015), would have defeated the purpose of sampling bears on their natural movements.
One strength of this study approach was the ability to examine samples from the same brown bear in the same or in different years. By contrast, other study approaches sampled brown bears when dead or when tagged, but not recovered (e.g., Hopkins et al. 2012; Van Daele et al. 2013). Our isotopic values tended to be consistent within years, but were not always consistent between years, suggesting shifts in diet owing to available prey or bear status. In some cases, the bear displayed changing isotope values, indicating a reduction in trophic position. Although this might seem inconsistent with the expected increase in access to salmon as brown bears grow larger, it is also possible that we took our initial samples from neonatal bears, whose signatures might exceed those of their mothers (Hobson et al. 2000). If so, the same bear might display less enrichment in subsequent years. However, the enrichment of offspring relative to their mothers is questioned (Jenkins et al. 2001), so this phenomenon may not fully explain some declines in trophic status. Further sampling, including more brown bears in more years, and also parent–offspring relationships, is needed to fully explain these anecdotal results. Regardless, the results revealed that Sockeye Salmon played an important role in brown bear diets in this system, consistent with other similar study systems. However, we also found that despite ready access to salmon and high rates of predation, many brown bears also fed to a large extent at lower trophic levels, and the reliance on salmon varied markedly among brown bears. This heterogeneity may reflect some combination of individuality in behavior (see also Mowat et al. [2017] for such evidence in brown bears and Bolnick et al. [2003] for a review of the ecology of individuals), age and status, variation in availability of alternative food sources, and differences in foraging before the arrival of salmon.
Supplemental Material
Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author for the article.
Table S1. List of individual brown bears Ursus arctos sampled in western Alaska, USA, from 2012 to 2015 included in the study of stable isotope ratios. For each bear, we provide the sex, foraging neighborhood (HHE = Happy, Hansen and Eagle creeks; BYW = Bear, Yako, and Whitefish creeks) where we collected the sample, and the means, standard deviations, and sample sizes for stable isotope values.
Found at DOI: https://doi.org/10.3996/JFWM-20-034.S1 (31 KB DOCX).
Table S2. Model selection using Akaike Information Criterion (AICc; Akaike 1974) to test for differences in stable isotope values (δ13C and δ15N) of hair samples from brown bears Ursus arctos visiting six salmon-bearing streams in southwestern Alaska, USA, from 2012 to 2015. We built models using the sex of the sampled bear (sex), the timing of sample collection (collection day-of-year), the stream neighborhood on which we collected the sample (neighbor region), and a random intercept for bear individual (bear indiv) as predictors. Listed are the top five models; models with ΔAICc < 2 are bolded. For each model, we calculated evidence ratios by comparing the log-likelihood of model in question (numerator) with the log-likelihood of the null model (denominator) per Burnham et al. (2011).
Found at DOI: https://doi.org/10.3996/JFWM-20-034.S1 (31 KB DOCX).
Table S3. Estimated regression parameters and standard errors (SEs) for candidate models (delta Akaike Information Criterion [ΔAIC] < 2) to test for differences in stable isotope values from brown bears Ursus arctos hair samples from six streams in Alaska, USA, from 2012 to 2015 (Akaike 1974).
Found at DOI: https://doi.org/10.3996/JFWM-20-034.S1 (31 KB DOCX).
Acknowledgments
We thank the College of the Environment, School of Aquatic and Fishery Sciences, School of Environmental and Forest Sciences, Richard and Lois Worthington Endowment, and the Mary Gates Endowment Research Scholarship for supporting this project and the University of Washington's Alaska Salmon Program, including the many people who were involved in collecting the hair samples. We are grateful to Lisette Waits and Jennifer Adams (Laboratory of Ecological, Evolutionary and Conservation Genetics, University of Idaho) for the DNA analysis and to Andrew Schauer for the stable isotope analyses at the University of Washington's IsoLab. We thank the Holtgrieve Ecosystem Ecology Laboratory (Seattle, Washington) for the space to prepare stable isotope samples and help in data interpretation. We also thank Joy Erlenbach and Charlie Robbins (Washington State University, Pullman) for sharing their knowledge of brown bear ecology and physiology, Gordon Holtgrieve for helpful comments on an earlier draft of the paper, and the Associate Editor and reviewers for exceptionally thorough and constructive comments on the submission.
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.
References
The findings and conclusions in this article are those of the author(s) and do not necessarily represent the views of the U.S. Fish and Wildlife Service.
Author notes
Citation: Ro H, Stern JH, Wirsing AJ, Quinn TP. 2021. Stable isotopes reveal variation in consumption of Pacific salmon by brown bears, despite ready access in small streams. Journal of Fish and Wildlife Management 12(1):40–49; e1944-687X. https://doi.org/10.3996/JFWM-20-034