OUP user menu

Evaluating Welfare of American Black Bears (Ursus americanus) Captured in Foot Snares and in Winter Dens

Roger A. Powell
DOI: http://dx.doi.org/10.1644/05-MAMM-A-014R2.1 1171-1177 First published online: 14 December 2005


Much research on wild mammals requires trapping, especially livetrapping, yet few methods used to capture wild mammals have been tested against an accepted standard for animal welfare and few data exist regarding physiological responses to capture. My coworkers and I livetrapped 208 American black bears (Ursus americanus) 356 times between May 1981 and August 2001 in the Pisgah National Forest in the Southern Appalachian Mountains by using Aldrich-type foot snares modified for bear safety with automobile hood springs and swivels spliced into cables. We outfitted most bears with transmitter collars and followed 18 bears to their winter dens. We outfitted 8 bears with transponder collars mounted with remotely dischargable darts loaded with anesthesia. We recorded the physical injuries of all bears handled and obtained 186 standard blood chemistry profiles from 112 bears. I compared the blood chemistry profiles of snared bears to profiles of bears in dens, to profiles for healthy, captive bears, and to profiles for wild bears that were collar-darted. Aldrich-type foot snares modified for bear safety, as we used them, and den handling met the accepted standard for trap injuries. Blood chemistry profiles indicated that bears captured in snares experienced high levels of physical exertion and were dehydrated. Blood chemistry parameters responsive to exertion increased with increasing injury scores.

Key words
  • animal welfare
  • bear
  • creatine kinase
  • lactate dehydrogenase
  • serum blood chemistry
  • snare
  • trap
  • Ursus americana

How animals are trapped and the impacts of trapping on animal populations are major societal concerns (Proulx and Barrett 1989) and traps that have minimal effects on animals minimize aberrant effects on behavior and ecological data. Yet, few traps have been tested against standards (Powell and Proulx 2003; Proulx 1999). Traps used in research should meet performance criteria that address state-of-the-art trapping technology and that optimize animal welfare conditions within the context of the research (Powell and Proulx 2003; Proulx 1999; Proulx and Barrett 1994).

Much research on wild mammals uses live, or restraining, traps. Very few restraining traps have been tested against standards for animal welfare (Proulx 1999). Tullar (1984) and Olsen et al. (1986) developed systems to score injuries caused by live traps. Olsen et al. (1986) proposed that each bruise, minor cut, or minor joint damage be scored between 5 and 50 points, depending on its extent; serious injuries each be scored >50 points; and severe injuries be scored >125 points. They proposed that an acceptable, threshold sum for all injuries is 50 points, above which an animal has received unacceptable injury. Similar scoring patterns have been proposed since (reviewed by Proulx 1999). Powell and Proulx (2003) adopted a criterion for assessing live traps that is consistent with the accepted standards for killing traps (Canadian General Standards Board 1984; Powell and Proulx 2003; Proulx and Barrett 1994): State-of-the-art live traps should with 95% confidence render ≥70% of animals caught with ≤50 points scored for physical injury.

In addition to injuries from traps, mammals respond to capture behaviorally and physiologically (Cattet et al. 2003; Kreeger et al. 1990; Proulx et al. 1993; Seddon et al. 1999; Warburton et al. 1999; White et al. 1991). Cattet et al. (2003) compared blood chemistries of grizzly bears (Ursus arctos) darted from helicopters to those of bears captured in foothold snares. Warburton et al. (1999) noted that blood chemistry of silver-gray brush-tailed possums (Trichosurus vulpecula) captured in soft-catch–type foothold traps differed from that of possums captured in cage traps and from captive animals. To date, no objective scoring system for live traps combines injuries with behavioral and physiological responses (Powell and Proulx 2003; Proulx 1999), at least in part because interpreting behavioral and physiological responses is not straightforward (Dawkins 1998). Perhaps because responses are confusing, researchers avoid publishing data. The consequence is that, with few data, understanding behavioral and physiological responses to capture is delayed.

Here I evaluate the physical injury and blood chemistry responses of American black bears (Ursus americanus) captured in Aldrich-type foot snares modified for bear safety and of bears handled in their winter dens. In addition, I compare the blood chemistries of captured bears to those of captive bears and free-ranging, wild bears drugged remotely via a dart mounted on a radiocollar. Specifically, I tested the following hypotheses. Hypothesis 1: Injury scores for bears captured in snares modified for bear safety and for bears handled in their winter dens meet the accepted standard for humane live capture. Hypothesis 2: Injury scores differ among groups of bears. Specifically, I derived the following hypotheses from field observations: Hypothesis 2a: Injury scores for snared juvenile bears are greater than those for adult bears. Hypothesis 2b: Injury scores do not differ between the sexes for snared bears. Hypothesis 2c: Injury scores for snared bears are greater than those for bears handled in their winter dens. Hypothesis 3: Physiological responses of bears captured in snares modified for bear safety indicate exertion and dehydration. Hypothesis 4: Physiological responses indicative of exertion and dehydration increase with increasing injury scores.

I want to stimulate other field biologists to subject their methods of handling animals to scrutiny and to report what they learn, whether their methods meet objective standards or not. We cannot improve our methods systematically until we know how they compare to objective standards.

Materials and Methods

Between May 1981 and August 2001, my coworkers and I livetrapped black bears for research related to ecology, animal behavior, and wildlife management (summarized by Kovach and Powell [2003], Mitchell and Powell [2004], and Powell et al. [1997]). The study area was the 220-km2 Pisgah Bear Sanctuary, located in the Pisgah National Forest approximately 35 km southwest of Asheville, North Carolina (35°28′N, 82°40′W) in the southern Blue Ridge Mountains of the southern Appalachians. Elevations ranged from 650 to 1,850 m. Annual rainfall often exceeded 200 cm and fog frequently enveloped high elevations. Hardwoods, such as oaks (Quercus), hickories (Carya), tulip poplar (Liriodendron tulipifera), and maples (Acer); and pines (Pinus) and hemlocks (Tsuga) were the most abundant trees.

We trapped bears predominantly by using Aldrich-type foot snares modified with automobile hood springs to provide cushioning for trapped bears (Johnson and Pelton 1980). When a trapped bear reached the end of its cable, which was bolted to a tree, cushioning from the hood spring prevented the bear from receiving an abrupt jolt. We took care to set snares where bears could not tangle their cables around small trees and shrubs, thereby negating the cushioning effect of the hood spring, and we took care to set snares where bears would be shaded during daytime. We set snares predominantly 100–300 m from backcountry trails, some snares were set more than 6 km from trailheads, and we baited snares with sardines, meat scraps, or old pastries. We checked traps each morning and handled bears promptly after all traps had been checked. However, when 2 or 3 bears were trapped on 1 day, a bear or bears sometimes remained in traps into the afternoon. When faced with multiple captures, we handled bears in an order that minimized total time for handling all bears with the constraint that bears that appeared stressed were handled first.

We immobilized bears with a mixture of ketamine hydrochoride and xylazine hydrochloride (approximately 200 mg of ketamine and 100 mg of xylazine per milliliter) administered with a jab stick or blowgun at a dosage of 1 ml/25 kg estimated weight. We monitored vital signs during handling (body temperature, pulse rate, and respiration rate); we cooled with water and ice the few bears that overheated and we warmed the few bears that became cooled by wrapping each in a space blanket, sometimes by wrapping a researcher in the blanket with the bear. We gave all bears ear tags and tattoos, took standard measurements, extracted an upper P1 tooth to estimate age by counting cementum annuli, and drew blood from the femoral vein. Bears were considered to be adults when >3 years of age; 2-year-old females who bred and produced cubs the following winter also were considered to be adults. We outfitted adult bears and some juvenile bears with transmitter collars, according to research goals at the time of capture.

In winter, when possible, my coworkers and I followed bears with transmitter collars to their dens, which were hollow trees, hollow logs, small caves, or open nests on the forest floor (Powell et al. 1997). We approached dens quietly and slowly, being careful not to disturb the bears or to flush them. We immobilized bears in accessible dens by using a blow dart or jab stick, changed their collars if necessary, and drew blood. We monitored vital signs as done during summer. In 1989 and 1990, we outfitted adult bears with 800-g transponder collars (3M, St Paul, Minnesota, and Wildlink, Brooklyn Park, Minnesota—Delgiudice et al. 1990) mounted with darts containing Telazol at a dosage of 5 mg/kg bear weight. I anesthetized bears wearing dart collars remotely when their transmitter signals indicated that they were mildly active and were confining their movements to a small area estimated to be ≤800 m away (Powell et al. 1997). Mildly active bears were unlikely to travel long distances before they could be found in the woods; they were never seen before their darts were discharged remotely.

All injuries to bears were recorded in the field. I scored injuries (Table 1) from data sheets by using the scoring system of Olsen et al. (1986) unless an injury could not be scored by that system, in which case I used the system of Hubert et al. (1996). Olsen et al. (1986) developed their scoring system by performing necropsies on kill-trapped animals, whereas we closely examined live animals for injuries. Because we would have missed some injuries obvious only from necropsy, my scores are biased low by an unknown, but presumably small, amount. I used the scoring system of Olsen et al. (1986) because it is objective, because it is well known, because it can be used to score injuries of live animals, and because researchers need an a priori set of measures to assess injuries to animals captured alive (Powell and Proulx 2003). My coworkers and I treated all injuries as well as possible under field conditions; our research team included at different times veterinarians, veterinary technicians, and veterinary students. Our research was approved by the Institutional Animal Care and Use Committees of North Carolina State University and Auburn University, and was in accordance with the principles and guidelines of the American Society of Mammalogists (Animal Care and Use Committee 1998) and of the Canadian Council on Animal Care (1984), (1993).

View this table:
Table 1

Scoring system for injuries received by bears during capture. Scores are those of Olsen et al. (1986) unless an injury was not scored by that system, in which case the score was that of Hubert et al. (1996) and is marked with an asterisk (*).

Edematous swelling and hemorrhage5
Avulsed claw (claw removed exposing pulp)*5
Cutaneous laceration < 2 cm length5
Permanent tooth fracture exposing pulp*10
Cutaneous laceration > 2 cm10
Tendon or ligament laceration20
Joint subluxation30
Joint luxation50
Compression fracture above or below carpus30
Simple fracture at or below carpus or tarsus50
Compound facture at or below carpus or tarsus75
Simple fracture above carpus or tarsus100
Compound fracture above carpus or tarsus200
Amputation of digits
1 digit50
2 digits100
3 digits150
4 or 5 digits200
Amputation above digits400

I used the general linear model of SAS (PROC GLM—SAS Institute Inc. 1999) to test for differences in injury scores among bears of different maturity and sex. Each capture was then scored binomially as being “Acceptable” (score ≤ 50) or not. The exact lower 95% confidence limit was calculated for the binomial probability that a bear–s injury score was “Acceptable.”

Blood chemistry profiles were generated using standard methods at Mission Memorial Hospital, Asheville, North Carolina (Table 2). for comparison, I obtained 2 sets of reference chemistry profiles. First, I obtained profiles (mean, SD, and sample size) for captive black bears through the International Species Inventory System (ISIS; Table 2). When the mean reference (ISIS) value for a blood parameter differed by 2 SDs from the mean value obtained for wild bears (either ISIS SD or wild SD), I considered the difference significant. I used this criterion because only summary statistics were available from ISIS (original data for each bear were not available). Because the distributions of some parameters are not normal, reporting medians and quartiles would be preferred were the ISIS data available in that format. Second, I used as reference the profiles for bears darted remotely by using darts mounted on transponder collars. I compared these blood chemistry profiles to those of bears captured in snares and handled in dens. These bears that were darted remotely in the wild should have had blood chemistry profiles representative of mildly active, wild bears.

View this table:
Table 2

Blood chemistry results. The 2nd column states whether a blood parameter was grouped as responsive to exertion (E) or dehydration (D) in multivariate analyses. The remaining columns list values for 112 bears captured 186 times in the Pisgah Bear Sanctuary during 1984–1996 and reference values for captive bears from the International Species Inventory System (ISIS). Both the exertion and the dehydration blood chemistry groups differed significantly in group values between adult and juvenile bears, and between both adult and juvenile bears and dart-collared bears.

Snared bears—allSnared bears—adultSnared bears—juvenileDenned bearsDart collarsISIS
Blood parametersGroup ± SDn ± SDn ± SDn ± SDn ± SDn ± SDn
Serum sodium (mmol/liter)D141 ± 5167141 ± 580141 ± 587139 ± 51136 ± 36139 ± 4119
Serum chloride (mmol/liter)D104 ± 6167104 ± 680104 ± 687101 ± 41103 ± 56104 ± 5101
Total serum protein (g/dl)D7.1 ± 0.61667.3 ± 0.6807 ± 0.5867.9 ± 0.477.6 ± 0.667.6 ± 0.6126
Albumin to globulin ratioD1.4 ± 0.41661.3 ± 0.3801.4 ± 0.4861.3 ± 0.370.7 ± 0.161.2
Total serum bilirubin (mg/dl)D0.2 ± 0.21660.2 ± 0.2800.3 ± 0.2860.2 ±060.3 ± 0.160.2 ± 0.299
Serum glucose (mg/dl)D, E170 ± 43167175 ± 4380165 ± 4487125 ± 56783 ± 356102 ± 37134
Serum albumin (g/dl)D, E4 ±0.51664.1 ± 0.5804 ± 0.5864.5 ± 0.573.2 ± 0.164.1 ± 0.5103
Serum CO2 (mmol/liter)E22 ± 416723 ± 48022 ± 58720 ± 5721 ± 5619 ± 539
Alkaline phosphatase (IU/liter)E54 ± 3516635 ± 227972 ± 368712 ± 5755 ± 22629 ± 19109
Alanine aminotransferase (IU/liter)E101 ± 9316796 ± 9980106 ± 878718 ± 8729 ± 56
Lactate dehydrogenase (IU/liter)E2,024 ± 2,0211661,693 ± 1,949802,332 ± 2,04886574 ± 3257598 ± 926644 ± 35645
Creatine kinase (IU/liter)E10,720 ± 17,8391629,299 ± 19,0087812,038 ± 16,68884328 ± 2727133 ± 346151 ± 12160
Serum cholesterol (mg/dl)E282 ± 50166271 ± 5080293 ± 4886265 ± 767296 ± 396268 ± 68107
Serum triglycerides (mg/dl)E268 ± 108166250 ± 10480284 ± 10986222 ± 867240 ± 776234 ± 7850
Amylase (IU/liter)E57.2 ± 49.24149.5 ± 36.71963.9 ± 57.92217 ± 13622 ± 1546
Serum potassium (mmol/liter)4.5 ± 0.51674.4 ± 0.5804.5 ± 0.4874.5 ± 0.474.9 ± 0.464.6 ± 0.4118
Blood urea nitrogen (mg/dl)14 ± 716712 ± 58016 ± 8877 ± 4721 ± 8620 ± 13121
Serum creatinine (mg/dl)1.4 ± 0.51671.6 ± 0.5801.2 ± 0.4872.8 ± 0.571.2 ± 0.362 ± 0.6118
Serum uric acid (mg/dl)1.6 ± 1.31661.3 ± 0.7801.8 ± 1.6862.6 ± 2.270.7 ± 0.361.4 ± 1114
Calcium (mg/dl)8.7 ± 0.81678.5 ± 0.9808.8 ± 0.8878.5 ± 178.2 ± 0.669.3 ± 0.6126
Serum phosphorus (mg/dl)4.6 ± 1.81663.8 ± 1.2805.3 ± 1.9863.7 ± 0.775 ± 0.865.6 ± 1118
Direct bilirubin (mg/dl)0.1 ± 0.21650.1 ± 0.1800.1 ± 0.2850 ± 060.2 ± 0.160 ± 0.134
Indirect bilirubin (mg/dl)0.1 ± 0.21640.1 ± 0.2790.2 ± 0.2850.2 ± 060.1 ± 0.160.2 ± 0.234
Gamma-glutamyltransferase (IU/liter)14 ± 916315 ± 97813 ± 8858 ± 5717 ± 56
Aspartate aminotransferase (IU/liter)589 ± 847166522 ± 83280652 ± 8618662 ± 16785 ± 156
Serum iron (mg/dl)128 ± 66152131 ± 5867125 ± 728520 ±448144 ± 675167 ± 9817

To avoid multiple cases of spurious significance when testing multiple hypotheses on many blood parameters, I used multivariate Kruskal–Wallis tests (multivariate analysis of variance [MANOVA] option PROC GLM on ranks—SAS Institute Inc. 1999). For these analyses, I grouped together those blood parameters expected to respond to exertion and dehydration (Kaneko 1989; Table 2). The exertion group included serum albumin, alanine aminotransferase, amylase, alkaline phosphatase, cholesterol, CO2, creatine kinase, glucose, and lactate dehydrogenase, whereas the dehydration group included albumin, albumin to globulin ratio, Cl, glucose, Na, total bilirubin, and total protein. Because some bears were captured more than once, I tested for an effect of individual bears and would have blocked tests had it been necessary. Where MANOVAs found significant variation for blood parameter groups, I tested individual blood chemistries by using analyses of variance (ANOVAs) to find those responsible for the significance. Because I performed many ANOVAs, I set statistical significance at P = 0.01 to be conservative and to avoid spurious acceptance of significance. Finally, I regressed the values for each blood parameter against the injury score for each bear captured by snare (PROC REG—SAS Institute Inc. 1999).

Hypotheses 1–4 were the null research hypotheses and were developed from field experience before data sheets were scored for injuries and before blood data were analyzed. For SAS and other non-Bayesian statistical software, however, hypotheses 1–4 are considered to be alternative hypotheses to a statistical null hypotheses assuming “no effect.” Biological null hypotheses, however, represent standard knowledge and are the norm against which alternative hypotheses should be tested. To test my biological null hypotheses by using conventional statistics, I used and report the probability of type II error for the tests of the statistical null hypotheses (F-values and degrees of freedom are given for the type I error of conventional statistics).


My research team and I livetrapped 208 black bears (125 males, 81 females, and 2 sex unknown) 366 times (340 in snares, 8 by dart collars, and 18 handled in their dens). Exertion and dehydration occurred during some captures and appeared to differ by capture method. Bears captured in snares often struggled vigorously to escape, whereas bears handled in dens and collar-darted did not struggle. Bears captured in snares were sometimes without water for hours, whereas collar-darted bears were normally hydrated.

Injury scores did not differ among snared bears of different sex and age classes (F = 0.64, d.f. = 3,388, P ≫ 0.05; Table 3). Injury scores for snared bears exceeded those for bears handled in their dens (F = 16.94, d.f. = 1, 390, P < 0.01; Table 3).

View this table:
Table 3

Injury scores for black bears of different age and maturity classes. Bears were captured in foothold snares modified for bear safety and were handled in their dens. To meet accepted standards for animal welfare, a method of capture should have an injury score ≤ 50 70% of the time with 95% confidence. The far right column shows the lower 95% confidence limit for the probability that injury score will be ≤50 for each method of capture as used in this study. Both foothold snares modified for bear safety and den handling met the criterion during this study.

Capture techniqueScoreNumber with score ≤ 50 (%)95% lower confidence limit that injury score will be ≤50 (must exceed 70% to be acceptable) (%)
SexMaturityX̄ ± SDN
FootholdAll bears8 ± 15340334 (97)94
snareFemaleAdult7 ± 984
FemaleJuvenile10 ± 1270
MaleAdult8 ± 861
MaleJuvenile10 ± 21125
DenAll bears1.4 ± 2.31818 (100)81

Aldrich-type foot snares modified for bear safety and handling bears in dens both met the standard of being 95% confident that injury scores will be ≤50 points for ≥70% of captures (Table 3). The 95% lower confidence limit was 94% of captures for modified Aldrich-type foot snares and 81% of captures for den handling (both well above 70%).

Between January 1984 and August 1996, my coworkers and I collected blood samples from 112 bears (66 males and 46 females) captured 186 times (171 times in snares, 8 by dart collars, and 7 in dens). Means (± SD) for all blood chemistries are given in Table 2. Blood chemistry profiles for male and female bears did not differ for either blood chemistry group (exertion or dehydration), nor did blood chemistry profiles differ for lactating compared to nonlactating females.

Individual values for amylase, creatine kinase, and lactate dehydrogenase were high for bears captured in snares compared to the ISIS values (Tables 2 and 4). In addition, albumin and uric acid values were low for collar-darted bears compared to ISIS values; and alkaline phosphatase, blood urea nitrogen, and serum phosphorus values were low for bears captured in their dens. However, levels for den capture were confounded by occurring only in winter, whereas ISIS values were taken during summer for active bears. The differences between ISIS and den blood chemistry profiles highlight the differences in renal function between active and denning bears (Nelson et al. 1973, 1983, 1984).

View this table:
Table 4

Blood chemistry parametersa with significant contrasts. Blood parameters known to respond to exertion or to dehydration differed as groups between adult and juvenile bears, and between bears trapped in snares and free-ranging bears wearing dart collars (multivariate analysis of variance). The individual parameters that contributed those significant differences (least significant differences means test, general linear model procedure of SAS—SAS Institute Inc. 1999) are listed below along with individual parameters that differed between snared bears and captive bears, between denned bears and free-ranging bears, and between free-ranging bears and captive bears. ISIS stands for blood profiles for captive bears provided by the International Species Inventory System.

Blood chemistry parameters with significant contrasts
ComparisonExertion parametersDehydration parametersAll parameters
Snared adults versus juvenilesAP, ALT, LDH, CK, TriTP
All snared bears versus dart collaredGlu, Alb, AP, ALT, LDH, CKGlu, Alb, A/G, TP
Snared adults versus dart collaredGlu, Alb, ALT, LDH, CKNa, Glu, Alb, A/G
Snared juveniles versus dart collaredGlu, Alb, ALT, LDH, CKNa, Glu, Alb, A/G
All snared bears versus ISIS Snared adults versus ISISLDH, CK, Amy LDH, CK
Snared juveniles versus ISISLDH, CK, Amy
Denned bears versus dart collarAlb, APAP
Denned versus ISISBUN, UA, P, AP
Dart collar versus ISISUA, Alb
  • a Alb: albumin; A/G: albumin to globulin ratio; ALT: alanine aminotransferase; Amy: amylase; AP: alkaline phosphatase; BUN: blood urea nitrogen; CK: creatine kinase; Glu: glucose; LDH: lactate dehydrogenase; Na: sodium; Tri: triglycerides; UA: uric acid.

The blood chemistry profiles in the exertion blood chemistry group differed among individual bears, differed by maturity (juveniles compared to adults), and differed between the sexes (for all P < 0.01). MANOVAs testing for effects of both individual bears and maturity and testing for effects of both sex and maturity showed that the significance found by testing for effects of individual bears and sex alone was actually caused by differences in maturity. Consequently, subsequent tests for differences were blocked for maturity effects.

Both exertion and dehydration profiles for bears captured in snares differed between adults and juveniles (exertion: F = 9.93, d.f. = 9,161, P < 0.01; dehydration: F = 4.01, d.f. = 7,163, P < 0.01; Tables 2 and 4), differed between snared bears and collar-darted bears (exertion: F = 4.65, d.f. = 9, 168, P < 0.01; dehydration: F = 7.33, d.f. = 7,170, P < 0.01; Tables 2 and 4), differed between snared adult bears and collar-darted bears (exertion: F = 9.93, d.f. = 9, 161, P < 0.01; dehydration: F = 4.01, d.f. = 7, 163, P < 0.01; Tables 2 and 4), and differed between snared juvenile bears and collar-darted bears (exertion: F = 5.23, d.f. = 9,87, P < 0.01; dehydration: F = 2.92, d.f. = 7, 89, P < 0.01; Tables 2 and 4). The statistical significances among these profiles were caused by individually significant differences in albumin, alanine aminotransferase, alkaline phosphatase, creatine phosphokinase, glucose, lactate dehydrogenase, and triglycerides for exertion; and albumin, albumin to globulin ratio, glucose, Na, and total protein for dehydration (Table 4).

Five blood parameters increased with increasing levels of injury: alanine aminotransferase, aspartate aminotransferase, creatine kinase, glucose, and lactate dehydrogenase (P < 0.05 for each). Of these significant parameters, alanine aminotransferase, creatine kinase, glucose, and lactate dehydrogenase were included in the exertion group.


I accept hypothesis 1. Modified Aldrich-type foot snares, as used by my research team, and handling bears in their winter dens met the accepted standard for injuries due to capture (Powell and Proulx 2003): we can be 95% confident that captured bears will have injury scores ≤ 50 points in ≥70% of captures. These statistical results are so strong that the potential bias (from scoring live bears, not doing necropsies) in my injury scores appears not to have affected the overall assessment of capture methods. Table 3 shows that acceptable injury scores should be obtained well in excess of 70% of captures.

I accept hypotheses 2b and 2c but cannot accept hypothesis 2a. Injury scores for bears captured in snares modified for bear safety were higher than those for bears handled in their winter dens. However, injury scores did not differ either between the sexes or for juvenile bears compared to adults. Although young bears struggled in snares more than adults, their small bodies and light weights apparently limited injuries due to struggling. Nonetheless, the struggling of juvenile bears did affect their blood profiles, leading to higher exertion and more dehydration than shown by adults (Table 2).

I accept hypothesis 3. Bears captured in snares had blood parameter values that differed from values for captive bears and from values for free-ranging, wild bears darted remotely (Tables 2 and 4). ISIS values for amylase, creatine phosphokinase, and lactate dehydrogenase differed from those of snared bears, consistent with higher activity levels and some dehydration by bears in snares. Snared bears differed significantly from collar-darted bears in their exertion and dehydration profiles. In addition to differences in creatine kinase and lactate dehydrogenase identified in comparisons with ISIS values, values for glucose, albumin, total protein, alkaline phosphatase, and alanine aminotransferase differed between snared bears and collar-darted bears.

Blood chemistry results were consistent with my anecdotal observations of captured bears: bears struggled in snares and some struggled vigorously. Bears that were collar-darted remotely were usually found in the woods as though they had simply lain down to sleep but in a position so as not to roll downslope. Bears captured in snares contrasted with collar-darted bears by having elevated values for alanine aminotransferase, alkaline phosphatase, creatine kinase, lactate dehydrogenase, glucose, albumin, and albumin to globulin ratio, reflecting exertion and dehydration (Tables 2 and 4). In addition, alanine aminotransferase, creatine kinase, and lactate dehydrogenase all increased with increasing injury scores, indicative of exertion. The high levels of aspartate aminotransferase with high injury scores could be an artifact of high levels of lactate dehydrogenase, which confounds the aspartate aminotransferase test (Kaneko 1989). Bears captured in snares also contrasted with collar-darted bears in level of dehydration by having elevated values for albumin, albumin to globulin ratio, total protein, and glucose (Tables 2 and 4).

I accept hypothesis 4. Creatine kinase, lactate dehydrogenase, glucose, and alanine aminotransferase all increased with increasing injury scores, indicating that increased exertion led to increased injury.

The exertion and dehydration profiles showed that adult bears captured in snares exerted themselves less and were less dehydrated than juveniles. This result is consistent with the anecdotal observations that young bears struggled more in snares, even though juvenile bears did not have significantly higher injury scores.

General discussion.—Blood chemistry profiles from ISIS for captive bears and for collar-darted, wild bears were similar, differing only in serum uric acid and albumin levels. Even though the sample size for collar-darted bears was small, use of the blood chemistry profiles of these bears for comparison appears sound, especially because these bears were wild, free-ranging, and naturally active.

Creatinine kinase, lactate dehydrogenase, and amylase were high in bears captured in snares, indicative of exertion compared to reference values from ISIS and to values for collar-darted bears. Warburton et al. (1999) and Cattet et al. (2003), respectively, noted elevated levels of these chemistries for brush-tailed possums and grizzly bears captured in foot-restraining traps. What level of these parameters might indicate excessive exertion? Wild black bears can be extremely active: they tear open logs, move large boulders, and climb trees. The effects of such exertion on blood profiles are unknown. The blood chemistry profiles for collar-darted bears are representative of mildly active bears; they are not representative for natural, strenuous activity. Before we can set criteria for physiological measures, we need physiological data for animals across the entire range of normal activities. Consequently, at present, no objective criteria exist to combine physiological measures with injury scores to evaluate whether traps are humane. Similarly, no behavioral data exist to develop objective criteria for humane trapping.

The blood chemistry levels reported here (Table 2) are similar to those reported elsewhere for black bears (Eubanks et al. 1976; Hellgren et al. 1989; Matula et al. 1980) and other bears (Cattet et al. 2003; Halloran and Pearson 1972), as is the lack of difference of blood profiles for males and females (Brannon 1985; Hellgren et al. 1989; Matula et al. 1980; Schroeder 1987). Consistent with these other studies, I did find that females had lower blood phosphorus levels, but not for the reason often given: that lactating females have high P demand (Brannon 1985). The lactating females I studied actually had higher P levels than did nonlactating adult females. The lack of significant difference between profiles of lactating and non-lactating females suggests that lactation does not place high physiological demands on bears in my study area. Given that lactation is energetically demanding in mammals (Kenagy et al. 1989, 1990; review by Moen 1973), bears in my study area appear to be in good condition with ample energy reserves and not subjected to food limitation. This suggestion is supported by data on food supplies and home ranges (Powell et al. 1997).

How foot snares are set is important (Mowat et al. 1994). Although we did not regularly note on data sheets whether a snared bear had tangled its cable around logs, saplings, or shrubs, anecdotal evidence noted in field journals indicated that bears were more likely to be injured when hood springs could not function well. Mowat et al. (1994) noted that tangled cables led to injuries for Canada lynxes (Lynx canadensis) trapped with foot snares. Johnson and Pelton (1980), Huber et al. (1996), and Graf et al. (1992) noted that a foot snare must be outfitted with something to cushion the strain when a bear struggles at the end of its cable. My research does not suggest carte blanc approval of foot snares for black bears but shows clearly that foot snares can be used in a manner that limits injuries of captured black bears to acceptable levels.

Foot snares, the traps most easily transported in the field and the easiest to spread across a backcountry study area, can meet accepted criteria for animal welfare. Nonetheless, researchers should always seek ways to reduce even more the effects they have on the animals they study. Pruss et al. (2002) outfitted neck snares for coyotes (Canis latrans) with diazepam tranquilizer tabs and noted reduced incidence of lacerations. Outfitting foot snares similarly might reduce exertion and dehydration after capture. If black bears struggle more at a particular time of day (e.g., after sunrise), traps might be checked before then to minimize struggling.

Field methods evolve by researchers dealing with conditions unique to each field site and to each study. Pawlina and Proulx (1999) recommended that traps and trapping methods be assessed only after an acclimatization period during which researchers become familiar with their methods.

Most important now is the need for researchers to evaluate their field methods with respect to objective criteria and to report the results. My coworkers and I worked hard over many years to minimize the impact of our trapping methods on the bears we studied, yet I did not know before doing the analyses reported here that our methods would meet accepted standards. To accept hypothesis 1 was rewarding, yet analyses of field methods should be reported whether field methods meet accepted standards or not. Without such evaluations, we do not know how well our methods stand up to accepted, objective standards, we cannot improve the treatment of animals captured for field research, we cannot reduce our impacts on their lives, and we chance having our methods bias our research results.


Graduate students G. Warburton, J. Zimmerman, M. Horner, M. Fritz, D. E. Seaman, J. Noel, A. Kovach, V. Sorensen, P. Mooreside, T. Langer, M. Reynolds, J. Sevin, J. Favreau, and L. Brongo, and Visiting Scientist F. Antonelli collected data. More than 30 undergraduate interns, technicians, and volunteers also assisted in data collection, as did personnel from the North Carolina Wildlife Resources Commission, and more than 300 Earthwatch volunteers. M. Stoskopf, M. Mitchell, and 4 anonymous reviewers made suggestions that significantly improved the manuscript. Financial and logistic support came from R. Bacon and K. Hailpern, the Bear Fund of the Wyoming Chapter of The Wildlife Society, D. Brown, J. Busse, Citibank Corp., the Conservation Fund of the Columbus (Ohio) Zoo, the Geraldine R. Dodge Foundation, Earthwatch/The Center for Field Research, Federal Aid in Wildlife Restoration Project W-57 administrated through the North Carolina Wildlife Resources Commission, Grand Valley State University McNair Scholars Program, International Association for Bear Research and Management, G. & D. King, McIntire Stennis funds, the National Geographic Society, the National Park Service, the National Rifle Association, the North Carolina Agricultural Research Service, North Carolina State University, 3M Co., the United States Department of Agriculture Forest Service, Wildlands Research Institute, Wil-Burt Corp., and Wildlink, Inc. Port Clyde and Stinson Canning Companies donated sardines.


  • Associate Editor was Floyd W. Weckerly.

Literature Cited

View Abstract