Printer Friendly

Effects of repeated captures on body mass and survival of dusky-footed woodrats in a California oak woodland.

Large amounts of mark-recapture data, accumulated over repeated trapping sessions, are generally required for precise and relatively unbiased estimates of demographic parameters (e.g., survival) of small mammals. Enumerative indices based on limited data from trapping sessions are often imprecise, always negatively biased, and unsuitable for studies of the dynamics of communities (Nichols and Pollock, 1983; Nichols, 1986; McKelvey and Pearson, 2001). While trapping effort varies among studies from 2-12 trapping occasions (also known as trap nights)/session, five occasions is the recommended minimum to make effective use of the power of prediction models in data analyses (Nichols and Coffman, 1999; Shanker, 2000). Generally speaking, more sampling effort and higher rates of recapture lead to better estimates of parameters, more rigorous results, and ultimately more valuable contributions to collective ecological knowledge.

Repeatedly capturing and handling small mammals, however, may invoke important ethical considerations for the use of animals in research and for the investment of research funds. Investigators have reported loss of mass and even direct trap-mortality of small mammals attributable to confinement in traps and handling. Results vary by species, cohort, reproductive status, live-trapping protocol (e.g., type of trap, type of bait, frequency of checks of traps), and environmental factors (Llewellyn, 1950; Bietz et al., 1977; Kaufman and Kaufman, 1994; Pearson et al., 2003). In temperate climates, greater losses were measured during drought conditions for male but not female North American deer mice (Peromyscus maniculatus; Kaufman and Kaufman, 1994) and with extremes in temperature for deer mice (Kaufman and Kaufman, 1994), hispid cotton rats (Sigmodon hispidus), and prairie voles (Microtus ochrogaster; Slade, 1991). In contrast, Suazo and Delong (2007) found nearly the opposite effect on oldfield deer mice (P. polionotus) in the warmer climate of Florida, with the least loss of mass occurring in summer, the hottest and wettest season.

Loss of mass during multiple trapping occasions might (Shanker, 2000; Pearson et al., 2003; Suazo et al., 2005) or might not (Kaufman and Kaufman, 1994; Pearson et al. 2003) be cumulative. The effect of rest between captures on loss of mass varies by species (Korn, 1987; Pearson et al., 2003) with the possibilities of lesser loss of mass, apparent complete recovery or gain of mass with longer intervals, or no apparent relationship. Slade and Iskjaer (1990), reporting mean losses for hispid cotton rats and prairie voles, found that the larger cotton rat (adults 100-150 g) lost proportionally less mass than did the smaller prairie vole (adults <50 g). Within species, lighter animals typically lost proportionally less mass than did heavier animals (Slade, 1991; Suazo et al., 2005). Similarly, juveniles lost less mass on average compared to the larger adults (Kaufman and Kaufman, 1994; Pearson et al., 2003).

Although examination of the literature leaves little doubt that small mammals exhibit short-term declines in mass due to confinement in traps and handling, fewer studies have investigated persistent long-term effects of live-trapping. However, those studies that have assessed longer-term effects suggest that repeated captures and associated declines in mass can have lasting impacts. For instance, Slade (1991) found that prairie voles that lost the most mass within a trapping session recouped the least amount of mass between sessions, suggesting that trap-induced losses in mass can have long-term effects on body condition. Pearson et al. (2003) reported that intervals of [less than or equal to] 6 days between consecutive captures did not lessen declines in mass for red-backed voles (Clethrionomys gapperi) or red-tailed chipmunks (Tamias ruficaudus). They also found that declines in mass associated with trap-mortality among P. maniculatus tended to be more severe than those observed for surviving animals, suggesting that declines in mass might have negatively impacted survival (Pearson et al., 2003). Moreover, Slade (1991) documented a negative association between loss in mass and monthly survival rates among prairie voles.

The dusky-footed woodrat (Neotoma fuscipes) is a medium-sized nocturnal murid rodent that occupies a variety of habitats throughout California, including chaparral (Lee, 1963), dense coniferous forests (Tevis, 1956), coastal sage-scrub (Horton and Wright, 1944), and oak woodlands (Ingles, 1995). Woodrats are semi-territorial (Innes et al., 2009), but the home ranges of oppositesex conspecifics can overlap considerably (Carraway and Verts, 1991), particularly during the breeding season (February-September; Vestal, 1938). Woodrats are well-known for their large stick houses (Linsdale and Tevis, 1951) that afford protection from extreme temperature, facilitate hoarding of food, and provide cover for numerous commensal species (Carraway and Verts, 1991). Although N. fuscipes is fairly well-studied and is considered an integral part of the biotic communities where it occurs, we are not aware of any study that examines either short-term or long-term impacts of trapping on this species.

As part of a long-term study on small mammals (demography, relationships with habitat, and response to management of land; e.g., Lee and Tietje, 2005), we used mark-recapture data collected during 2002-2006 to quantify the short-term effects of repeated captures on woodrats in a relatively mild, two-season climate (cool wet winter versus hot dry summer) to provide a point of comparison with the results and implications of other studies. We also used the data to investigate long-term effects on survival of woodrats. In accordance with the literature, we expected short-term, trap-induced declines in mass among woodrats and anticipated that change in mass might be influenced by number of captures and the age and sex of captured animals. We also reasoned that cumulative losses in mass due to repeated captures might have significant long-term consequences for the well-being of an animal and, therefore, expected a negative association between the number of captures during a trapping session and survival to the next trapping session (ca. 5 months).

MATERIALS AND METHODS--We trapped woodrats each fall and spring at the United States Army National Guard Post, Camp Roberts, ca. 18 km north of Paso Robles, San Luis Obispo County, California. The study area received only minimal use by military personnel. Habitat was an oak-woodland matrix of grassland, chaparral, and open-to-dense woodland. The canopy was comprised of blue oak (Quercus douglasii) mixed with coast live oak (Q. aggrifolia) in the more densely wooded areas. Patches of shrubs (Heteromeles arbutifolia, Toxicodendron diversilobum, Ceanothus, Rhamnus, Arctostaphylos) were common, especially in the mixed woodland areas (Tietje et al., 1997). Ground cover was mostly wild oats (Avena) and brome grasses (Bromus), mixed with native forbs in the more dense areas. During trapping sessions, daily minimum and maximum temperatures averaged 3-8[degrees]C and 21-27[degrees]C, respectively, and precipitation ranged from nearly 0-3 mm/session (University of California IPM online,

We trapped small mammals for 3 consecutive days/trapping session (spring and fall 2002-2006) on 22 study plots, each 1.1-ha with an 8-X-8 grid with 15-m spacing between grid intersections. Trapping sessions were ca. 5 months apart, but the starting date of sessions varied somewhat from year to year. Within each of the 22 plots, a single Sherman live-trap (7.6-x-9.5x-30.5 cm nonperforated XLK aluminum folding trap; H. B. Sherman Traps, Inc.[R], Tallahassee, Florida) was set within 2 m of each grid intersection (64 traps/plot), covered with grass and duff from the immediate vicinity to help insulate captured animals from potentially cool overnight temperature or heating by sunshine in the early morning, and baited with 25-30 g of rolled corn, oats, and barley laced with molasses. Traps were not prebaited. We checked traps each morning beginning at ca. 0730 h. Because woodrats are nocturnal, and thus likely to enter traps only at night, we estimate that animals spent <12 h/ capture confined in a trap. We ear-tagged each captured woodrat for individual identification and recorded species, sex, age (juvenile or adult pelage), reproductive status (whether females were visibly pregnant or lactating), and mass (to the nearest gram). Mass was measured with a 300-g Pesola[R] spring scale (Model 40300; -0.3% accuracy; Pesola AG, Baar, Switzerland) by suspending the woodrat from the eye-clip of the scale clamped at the base of its tail. Bait was replenished as needed such that the amount in the trap sometimes exceeded the initial amount. We released animals at site of capture.

One person (WDT) processed approximately a third of the captured animals during 2002-2004 and all of them during 2005-2006, likely increasing the repeatability and consistency of measurements of mass taken within and among trapping sessions. During trapping sessions, wind was negligible and not a factor in the accuracy and consistency of measurements. Rainfall was uncommon during trapping sessions; however, we did not weigh animals noticeably damp from overnight rainfall. Trapping met the guidelines of the American Society of Mammalogists (Sikes et al., 2011) and was approved by the Animal Care and Use Committee of the University of California, Berkeley.

We defined initial mass as the body mass in grams when a woodrat was first captured in a trapping session, recognizing that this represents an initial response of unknown magnitude and direction. We calculated percentage of change in mass for subsequent captures relative to this initial measurement. We considered age and sex separately but, due to the relatively small sample of juveniles (n = 198), also combined the two factors into a single variable, age-sex, with the following four levels: nonreproductive adult female (AF); pregnant or lactating adult female (AFPL); adult male (AM); juvenile (J). We tallied the number of captures per session for each individual and defined the following four categories of capture: captured once; captured twice on consecutive days; captured twice with a rest day between; captured on all 3 days. Analyses were performed using R version 2.13.1 (R Development Core Team, R Foundation for Statistical Computing, Vienna, Austria; ISBN 3-900051-07-0, URL Results are presented as means [+ or -] SE.

Due to the relatively long span of time (ca. 5 months) between sessions, for purposes of change in mass, we considered captures and measurements of the same individual in multiple sessions (=observations) to be statistically independent. However, because some animals (e.g., woodrats occupying a house close to a trapping station) might be more likely than others to be captured multiple times within a session and in multiple sessions, when comparing categories of capture among age and sex classes, we randomly selected one observation (i.e., any one of multiple sessions in which an individual was captured one, two, or three times). We used categorical data analyses (chi-square) to examine differences in the number and pattern of captures by age, sex, reproductive status, season, and month.

Because we did not know whether several relatively large ([greater than or equal to] 15 g) losses in mass between consecutive captures of reproductive females were due to parturition, to altered energetics during the trapping session (e.g., Mattingly and McClure, 1985), or to some other factor, we did not examine change in mass among reproductive females. We modeled variation in percentage of change in mass of individual woodrats between captures using fixed-effect analysis of variance (ANOVA). We started with the full model including age and sex, category of capture, and spatial and temporal variables (plot, season, year, and their first-order interactions) and used a backward-stepwise-elimination process to determine the highest-ranked reduced model using Akaike's Information Criterion (AIC) values. We repeated this procedure twice, first including age and sex, then substituting the combined age-sex variable (AF, AM, or J) for age and sex. We used Tukey-adjusted-multiplecomparison tests to identify significant effects at [alpha] = 0.05. For animals caught on 3 consecutive days, we used a paired t-test to compare change in mass between capture days 1 and 2 (first interval) and capture days 2 and 3 (second interval).

To quantify the effects of repeated captures on apparent survival (u) and recapture rates (p) of woodrats, we developed a set of Cormack-Jolly-Seber live-recapture models (Cormack, 1964; Jolly, 1965; Seber, 1965) using Program MARK (White and Burnham, 1999). We started with a general model where survival and recapture differed between the sexes and varied among trapping sessions. This model also included interactions between sex and session (i.e., [[phi]*t] [*t]). First, we assessed the fit of this global model by comparing the model deviance with the deviances from 1,000 simulations generated within MARK with the bootstrap goodness-of-fit procedure. We estimated the variance inflation factor (c) by dividing the observed deviance by the mean deviance of the 1,000 bootstrap simulations. Comparison of the global model deviance with the deviances of the simulated data suggested that our model adequately fit the observed data for woodrats (P = 0.26). We did not apply a variance inflation factor because P was sufficiently large (Burnham and Anderson, 2002) and because the estimated variance inflation factor was approximately equal to 1 (c = 1.05), suggesting a negligible amount of overdispersion in the data.

Next, we developed 15 reduced models representing various combinations of sex and time-varying effects on apparent survival and recapture. We selected our top model from this set using corrected AIC ([AI.sub.c]) values and then developed three additional candidate models by adding a covariate for number of captures per session (0-3) to the highest-ranked model. We used number of captures rather than change in mass because we were not able to quantify change in mass for animals that were not captured during a session or for animals trapped only once within a session. We modeled number of captures as an additive effect on [phi], p, and both, assuming a linear relationship and a constant effect over time (i.e., the strength of the capture effect did not vary among sessions). To avoid over-fitting the data, we did not consider interactions between number of captures and other variables. Finally, we used Program CONTRAST (Sauer and Hines, 1989) to test for significant differences in rates of apparent survival and recapture.

RESULTS--During 2002-2006, we recorded 2,864 observations (i.e., an individual captured one to three times within a trapping session) representing 4,705 total captures of 1,824 individual woodrats. Of the 2,864 observations, nearly all (93%, n = 2,666) were animals with adult pelage, and 53% (n = 1,514) were female (Table 1). The total captures were split nearly equally between spring (n = 2,343) and fall (n = 2,362), and approximately a third of the total occurred on each of the 3 days within a session (n = 1,476, 1,639, and 1,590, respectively). Approximately 35% of individuals (n = 639) were captured in more than one session. Within trapping sessions (n = 10), each individual was captured on average ca. 1.6 [+ or -] 0.01 times, with 52% (n = 1,497) captured only once, 31% (n = 895) captured twice, and 16% (n = 472) captured three times (Table 1). Of those animals caught twice, 74% (n = 664) were captured on consecutive days and 26% (n = 231) were caught on days 1 and 3.

Rates of capture within sessions (i.e., number of captures per session) varied somewhat by year, month, season, age and sex, and mass of the animal. Average capture rates ranged from 1.6-1.7 by year and from 1.51.9 by month (March, April, May, June, September, October, November). Little trapping was conducted in March and, therefore, the few (n = 14) captures of woodrats limit any inferences made involving rates of capture in that month. Overall, there was no difference in the number of captures per trapping session in spring compared to fall (1.62 [+ or -] 0.03 in spring versus 1.59 [+ or -] 0.02 in fall, [chi square] = 3.0, df = 2, P = 0.22). Compared to adults, juveniles were more likely to be captured only once within a session ([chi square] = 21.4, df = 2, P < 0.0001). Compared to males, females were more likely to be captured three times within a session ([chi square] = 9.2, df = 2, P = 0.01). Factoring in reproductive status, the number of captures within a session differed by age and sex ([chi square] = 46.0, df = 6, P < 0.0001) such that reproductive females were twice as likely to be captured three times as compared to other animals.

We recorded body mass for 4,279 (91%) of the 4,705 captures. We had sufficient data to calculate changes in mass for 1,193 observations. Initial mass and mean changes in mass by age and sex are presented in Table 2. Of the 931 observations of nonreproductive animals, 93% (n = 866) lost mass (mean = -5.2 [+ or -] 0.10%, range of - 19.5 to -0.33%) and the remaining 7% (n = 65) either gained mass or did not change in mass (mean = 2.1 [+ or -] 0.44%, range of 0-20.6%). For woodrats caught and weighed on 3 consecutive days, change in mass during the first interval (i.e., days 1 and 2) was slightly greater than during the second interval (i.e., days 2 and 3; first interval mean = -3.9 [+ or -] 0.1%; second interval mean = -3.2 [+ or -] 0.2%; paired-t = -3.42,df = 277, P = 0.0007). Within all age, sex, and weight groupings, the pattern of more loss in mass during the first period held true, and, within each group, the effect size (i.e., the difference in the percentage of mass lost between the first and second interval) was about equally large.

Both of our initial models examining different combinations of age and sex on loss of mass reduced to the same final model (Table 3). This model explained 31% of the variation in percentage of change in mass ([F.sub.51.879] = 7.75, P < 0.0001) and retained six factors: category of capture ([F.sub.2,928] = 114.88, P < 0.0001); season ([F.sub.1,929] = 4.30, P = 0.04); year ([F.sub.4,926] = 2.22, P = 0.07); plot ([F.sub.21,909] = 2.31, P < 0.0001); plot x season ([F.sub.19,911] = 3.52, P < 0.0001); season x year ([F.sub.4,926] = 9.18, P < 0.0001). Category of capture accounted for 18% of the variation in percentage of change in mass and 58% of the sums of squares of the model; whereas all spatial and temporal factors combined accounted for 13 and 42%, respectively. Change in mass differed by category (range of -3.5 to -6.8%), such that animals trapped three times lost more mass than did animals trapped on 2 consecutive days or twice with a day of rest; the latter two categories were not significantly different (Table 3). Animals captured in fall lost marginally more mass on average than did animals captured in spring (-4.8 [+ or -] 0.1% in fall versus -4.5 [+ or -] 0.2% in spring, P = 0.04; Table 3). Annual differences in change in mass were not significant (0.09 [less than or equal to] P < 0.93). The plot-level effect was driven by a single plot where mean loss in mass was significantly greater relative to six other plots (range of -1.9 to -2.9%); the remaining 224 pairwise comparisons were not significant.

The significant plot x season interaction accounted for about 17% of the sums of squares of the model, although relatively few pairwise comparisons were significant (38 of 946 possible comparisons). Season entered the model again through its interaction with year, ranging from the greatest loss of mass (-6.2 [+ or -] 3.6%) in spring 2002 to the least loss (-2.6 [+ or -] 0.5%) in spring 2003 (Table 3), and accounted for about 9% of the sums of squares of the model; 10 of 45 possible pairwise comparisons were significant. Neither the plot nor the season effects were consistent among years.

Of the 16 initial candidate models, the global model received the most support ([AIC.sub.c] = 4042.22, [w.sub.i] = 0.85, [DELTA][AIC.sub.c] range of 3.62-109.82). Of the final set of 19 candidate models, the top model included effects for sex and trapping session, an interaction between sex and trapping session, and a capture effect on survival ([AIC.sub.c] = 4032.14, [w.sub.i] = 0.66). The coefficient estimate for the capture effect on survival was positive and significant ([beta] = 0.406; 95% CI of 0.140, 0.673), indicating that animals captured more times within a session had higher survival to the next session. One other model could be considered competitive ([w.sub.i] = 0.24, [DELTA][AIC.sub.c] = 2.06). This model also included a capture effect on p, but the coefficient estimate was not statistically significant ([beta] = -0.004; 95% CI of -0.190, 0.181) and the effect explained a negligible amount (<0.01) of additional deviance. Consequently, we did not average coefficient estimates for the two models. Because survival and recapture were time-dependent in our highest-ranked model, estimates of [phi] and p for the final interval are inseparable (Lebreton et al., 1992); the results presented below only consider individually estimable parameters.

Rates of recapture (p) varied significantly over the course of the study for both sexes (female range of 0.56 [+ or -] 0.07 to 0.90 [+ or -] 0.05, [chi square] = 37.80, df = 7, P < 0.0001; male range of 0.43 [+ or -] 0.11 to 0.81 [+ or -] 0.09, [chi square] = 37.17, df = 7, P < 0.0001; Fig. 1a), and for all animals combined ([chi square] = 25.54, df = 7, P = 0.0006). Overall (i.e., all trapping sessions combined), females were more likely than males to be recaptured (average p = 0.71 [+ or -] 0.05 for females versus 0.56 [+ or -] 0.07 for males; [chi square] = 15.15, d.f. = 1, P = 0.0001). The difference between sexes was driven by rates of capture during the spring; females were more likely than males to be recaptured during spring sessions ([chi square] = 44.24, df = 1, P < 0.0001) but not during fall sessions ([chi square] = 0.26, df= 1, P = 0.61). Rates of recapture for all animals combined did not differ significantly between spring and fall sessions ([chi square] = 1.37, df = 1, P = 0.24). However, rates of recapture for females were higher during spring sessions than during fall sessions ([chi square] = 27.78, df = 1, P< 0.0001), whereas males were more likely to be recaptured in the fall than in the spring ([chi square] = 4.42, df = 1, P = 0.04; Fig. 1a).

Apparent survival (u) differed among trapping sessions for females (range of 0.58 [chi square] 0.05 to 0.80 [chi square] 0.05; [chi square] = 21.48, d.f. = 7, P = 0.003; Fig. 1b); differences in survival for males were marginally significant (range of 0.39 [+ or -] 0.09 to 0.66 [+ or -] 0.08; [chi square] = 13.75, df = 7, P = 0.06; Fig. 1b). Survival was similar between spring and fall for both sexes (females, [chi square] = 1.27, df= 1, P= 0.26; males, [chi square] = 0.07, df = 1, P = 0.80). Overall, female woodrats had higher survival than did males (average [phi] = 0.63 [+ or -] 0.03 for females versus 0.54 [+ or -] 0.04 for males; [chi square] = 5.59, df = 1, P = 0.02). Overall survival differed significantly for animals captured once, twice, or three times during trapping sessions (Fig. 2); this was true for females ([chi square] = 10.17, df = 2, P = 0.006), males ([chi square] = 6.29, df = 2, P = 0.04), and all animals combined ([chi square] = 15.10, df = 2, P= 0.0005). Woodrats captured once had lower survival than animals captured either two or three times (females, [chi square] = 10.16, df = 1, P = 0.001; males, [chi square] = 6.24, df = 1, P = 0.01; all animals, [chi square] = 15.04, df = 1, P = 0.0001), but there was no difference in survival between the latter two categories for females ([chi square] = 1.27, df = 1, P = 0.26), males ([chi square] = 0.89, df = 1, P = 0.35), or all animals combined (v2 = 2.03, df = 1, P = 0.15).

DISCUSSION--Contrary to our expectations, survival to the following trapping session was positively associated with the number of captures within a session. This result is surprising and should not be interpreted as a beneficial effect of trapping. Given that survival did not differ significantly between animals captured two or three times within a session, this effect is probably driven by lower survival among animals that were captured only once within a session (Fig. 2). We observed that juveniles, smaller woodrats, and the largest (presumably oldest) woodrats were more likely than other animals to be captured only once within a session. Body mass is generally correlated with survival among small mammals (e.g., Myers and Master, 1983; Sauer and Slade, 1986), and younger, smaller individuals might be more susceptible to mortality due to starvation, exposure, predation, or other sources. Lee and Tietje (2005) determined that juvenile woodrats at Camp Roberts had lower survival than adults 1-2 years old. They further reported that survival of woodrats declines markedly after ca. 2 years of age. It, therefore, seems plausible that lower survival among animals captured only once is simply an artifact of generally lower survival among younger and older woodrats.

Our results do not support the hypothesis that repeated captures of woodrats up to 3 consecutive days have a cumulative negative effect on survival, at least when intervals between trapping sessions are relatively long and offer ample time for individuals to recover from trap-induced loss in mass. Nevertheless, we note that persistent trap-induced declines in mass can depress longer-term survival when intervals between trapping sessions are shorter (Slade, 1991) and that the cumulative impact of repeated captures can result in severe loss in mass with the possibility of trap-mortality within longer sessions (Pearson et al., 2003). However, because we did not investigate the relationship between survival and change in mass per se, our results are not directly comparable to the findings of Slade (1991) and Pearson et al. (2003).

Overall, and especially during spring sessions, female woodrats were more likely to be recaptured than males. Higher probability of recapture among females might be because females have smaller home ranges than do males (Carraway and Verts, 1991), particularly during the main breeding season (Cranford, 1977), placing them in relatively close proximity to traps. Additionally, reproductive females might prioritize foraging to meet the energetic demands of pregnancy and lactation, and, therefore, be more likely to enter a trap to consume palatable bait. Thus, reproductive female woodrats are potentially the most vulnerable to the effects of capture, especially because they were more likely to be captured multiple times within a session than other age and sex classes. Confinement in traps might limit crucial foraging time and keep females apart from neonates dependent on frequent nursing. These potential effects of trapping should be investigated.

As predicted, woodrats exhibited short-term losses in mass associated with repeated captures. The losses in mass we observed are comparable to those documented elsewhere in the literature on wild small mammals. Korn (1987) reported that daily fluctuations in mass of 5-10% due to variations in content of the gut, hydration, and other factors are common, and these fluctuations are similar in magnitude to trap-induced losses observed in our study and references herein. Losses in mass were significantly greater for animals captured three times than for animals captured twice, indicating that changes in body mass were cumulative. This finding is not uncommon and has been reported for a variety of other species of small mammals (e.g., Korn, 1987; Pearson et al., 2003; Suazo et al., 2005). Losses in mass for animals trapped on 3 consecutive days were greater during the first interval than during the second interval. However, because we were not able to determine change in mass due to initial capture, we are unsure if this was true for all three intervals.

Differences in change in mass between spring and fall sessions were small and only marginally significant (P = 0.04; Table 3). Other investigators (e.g., Slade, 1991; Kaufman and Kaufman, 1994; Suazo and Delong, 2007) have reported stronger seasonal effects than those observed in our study. These differences are likely due to the comparatively mild environmental conditions during spring and fall in our study area. Kaufman and Kaufman (1994) reported highly significant declines in mass coincident with dry years and elevated temperatures in their study of P. maniculatus in Kansas, but animals in our study were not subjected to such extreme variation in environmental conditions. Maximum daily temperatures during our study (21-27[degrees]C) were comparable to the thermal neutral zone of the dusky-footed woodrat (20-25[degrees]C; Carraway and Verts, 1991), and animals did not experience the prolonged drought conditions described by Kaufman and Kaufman (1994). Therefore, we suspect that loss of water was not a significant cause of loss in mass in our study, but it might be a problem for animals captured during hot dry summers in California.

However, daily minimum temperatures during our trapping sessions were well below the thermal neutral zone for woodrats, and the minor differences in loss in mass between spring and fall sessions might have been driven by thermogenesis due to colder weather in the fall. These results are consistent with the greater consumption of energy required among terrestrial vertebrates to maintain thermal equilibrium at lower temperatures. Slade (1991) found that, even when traps were provisioned with bedding material, declines in mass were generally greatest during the coldest season for hispid cotton rats and especially for prairie voles. Although insulating our traps by covering them with grass and duff might have somewhat ameliorated the effects of colder weather during the fall, we note that woodrats in our study did not experience the extreme temperatures (< 0[degrees]C) reported by Slade (1991). However, ambient temperatures can drop below freezing in our study area during the winter, and woodrats trapped during winter sessions might exhibit greater declines in mass than those observed during the spring and fall in our study.

Although the declines in mass reported here were all considerably less than the >30% loss in mass due to severe starvation described by Robin et al. (2008), we note that losses were cumulative and might exceed this threshold during trapping sessions longer than 3 days. Additionally, we do not know the extent to which losses in mass might otherwise influence individual fecundity or fitness and thereby alter demographics. The effects of confinement in traps on the ability of woodrats to maintain territories or social status, compete with conspecifics, secure resources, find mates, or raise young are not well-studied and warrant further investigation. All of these possibilities have important consequences for individuals and might ultimately influence demography or the dynamics of populations. Consequently, we urge investigators to carefully consider the short-term effects of live-trapping on the well-being of individual animals as well as the long-term demographic impacts of trap-related effects on populations of rodents. We recommend further study to elucidate these potential adverse effects along with continued vigilance monitoring effects of the investigator on small mammals.

The University of California, Division of Agriculture and Natural Resources, Integrated Hardwood Range Management Program supported this research. We thank the California Army National Guard, Camp Roberts, for permitting access to study sites. The University of California Cooperative Extension Office, County of San Luis Obispo, provided logistical support. Anonymous reviewers, to whom we are grateful, provided helpful comments. M. Grajales-Hall and E. Moscoso prepared the Spanish translation of the abstract.


BIETZ, B. F., P. H. WHITNEY, AND P. K. ANDERSON. 1977. Weight loss of Microtus pennsylvanicus as a result of trap confinement. Canadian Journal of Zoology 55:426-429.

BURNHAM, K. P., AND D. R. ANDERSON. 2002. Model selection and multimodel inference: a practical information-theoretic approach. Second edition. Springer-Verlag, New York.

CARRAWAY, L. N., AND B. J. VERTS. 1991. Neotoma fuscipes. Mammalian Species 386:1-10.

CORMACK, R. M. 1964. Estimates of survival from the sighting of marked animals. Biometrika 51:429-438.

CRANFORD, J. A. 1977. Home range and habitat utilization by Neotoma fuscipes as determined by radiotelemetry. Journal of Mammalogy 58:165-172.

HORTON, J. S., AND J. T. WRIGHT. 1944. The wood rat as an ecological factor in southern California watersheds. Ecology 25:341-351.

INGLES, L. G. 1995. Mammals of the Pacific states. Stanford University Press, Stanford, California.

INNES, R. J., D. H. VAN VUREN, D. A. KELT, J. A. WILSON, AND M. L. JOHNSON. 2009. Spatial organization of dusky-footed woodrats (Neotoma fuscipes). Journal of Mammalogy 90:811-818.

JOLLY, G. M. 1965. Explicit estimates from capture-recapture data with both death and immigration-stochastic model. Biometrika 52:225-247.

KAUFMAN, G. A., AND D. W. KAUFMAN. 1994. Changes in body mass related to capture in the prairie deer mouse (Peromyscus maniculatus). Journal of Mammalogy 75:681-691.

KORN, H. 1987. Effects of live-trapping and toe-clipping on body weight of European and African rodent species. Oecologia 71:597-600.

LEBRETON, J. D., K. P. BURNHAM, J. CLOBERT, AND D. R. ANDERSON. 1992. Modeling survival and testing biological hypotheses using marked animals--a unified approach with case-studies. Ecological Monographs 62:67-118.

LEE, A. K. 1963. The adaptations to arid environments in woodrats of the genus Neotoma. University of California Publications in Zoology 64:57-96.

LEE, D. E., AND W. D. TIETJE. 2005. Dusky-footed woodrat demography and prescribed fire in a California oak woodland. Journal of Wildlife Management 69:1,211-1,220.

LINSDALE, J. M., AND L. P. TEVIS, JR. 1951. The dusky-footed wood rat. University of California Press, Berkeley.

LLEWELLYN, L. M. 1950. Reduction of mortality in live-trapping mice. Journal of Wildlife Management 14:84-85.

MATTINGLY, D. K., AND P. A. MCCLURE. 1985. Energy allocation during lactation in cotton rats (Sigmodon hispidus) on a restricted diet. Ecology 66:928-937.

MCKELVEY, K. S., AND D. E. PEARSON. 2001. Population estimation with sparse data: the role of estimators versus indices revisited. Canadian Journal of Zoology 79:1,754-1,765.

MYERS, P., AND L. L. MASTER. 1983. Reproduction by Peromyscus maniculatus: size and compromise. Journal of Mammalogy 64:1-18.

NICHOLS, J. D. 1986. On the use of enumeration estimators for interspecific comparisons, with comments on a 'trappability' estimator. Journal of Mammalogy 67:590-593.

NICHOLS, J. D., AND C. J. COFFMAN. 1999. Demographic parameter estimation for experimental landscape studies on small mammal populations. Pages 287-309 in Landscape ecology of small mammals (G. W. Barrett and J. D. Peles, editors). Springer-Verlag New York, Inc., New York.

NICHOLS, J. D., AND K. H. POLLOCK. 1983. Estimation methodology in contemporary small mammal capture-recapture studies. Journal of Mammalogy 64:253-260.

PEARSON, E., Y. K. ORTEGA, AND L. F. RUGGIERO. 2003. Trap-induced mass declines in small mammals: mass as a population index. Journal of Wildlife Management 67:684-691.

ROBIN, J. P., F. DECROCK, G. HERZBERG, E. MIOSKOWSKI, Y. LE MAHO, A. BACH, AND R. GROSCOLAS. 2008. Restoration of body energy reserves during refeeding in rats is dependent on both the intensity of energy restriction and the metabolic status at the onset of refeeding. Journal of Nutrition 138:861-866.

SAUER, J. R., AND J. E. HINES. 1989. Testing for differences in survival or recovery rates using program CONTRAST. Wildlife Society Bulletin 17:549-550.

SAUER, J. R., AND N. A. SLADE. 1986. Field-determined growth rates of prairie voles (Microtus ochrogaster): observed patterns and environmental influences. Journal of Mammalogy 67:61-68.

SEBER, G. A. F. 1965. A note on the multiple recapture census. Biometrika 52:249-259.

SHANKER, K. 2000. Small mammal trapping in tropical montane forests of the Upper Nilgiris, southern India: an evaluation of capture-recapture models in estimating abundance. Journal of Bioscience 25:99-111.

SIKES, R. S., W. L. GANNON, and THE ANIMAL CARE AND USE COMMITTEE OF THE AMERICAN SOCIETY OF MAMMALOGISTS. 2011. Guidelines of the American Society of Mammalogists for the use of wild mammals in research. Journal of Mammalogy 92:235-253.

SLADE, N. A. 1991. Loss of body mass associated with capture of Sigmodon and Microtus from northeastern Kansas. Journal of Mammalogy 72:171-176.

SLADE, N. A., AND C. ISKJAER. 1990. Daily variation in body mass of free-living rodents and its significance for mass-based population dynamics. Journal of Mammalogy 71:357-363.

SUAZO, A. A., AND A. T. DELONG. 2007. Responses of old-field mice (Peromyscus polionotus) to consecutive days of live trapping. American Midland Naturalist 158: 395-402.

SUAZO, A. A., A. T. DELONG, A. A. BARD, AND D. M. ODDY. 2005. Repeated capture of beach mice (Peromyscus polionotus phasma and P. p. niveiventris) reduces body mass. Journal of Mammalogy 86:520-523.

TEVIS, L., JR. 1956. Responses of small mammal populations to logging of Douglas-fir. Journal of Mammalogy 37:189-196.

TIETJE, W. D., J. K. VREELAND, N. R. SIEPEL, AND J. L. DOCKTER. 1997. Relative abundance and habitat associations of vertebrates in oak woodlands in coastal-central California. Pages 391-400 in Proceedings of the symposium on oak woodlands: ecology, management and urban interface issues (N. H. Pillsbury, J. Verner, and W. D. Tietje, technical coordinators). United States Department of Agriculture, Forest Service, General Technical Report PSW-GTR-160:1-738.

VESTAL, E. H. 1938. Biotic relations of the wood rat (Neotoma fuscipes) in the Berkeley Hills. Journal of Mammalogy 19:1-36.

WHITE, G. C., AND K. P. BURNHAM. 1999. Program MARK: survival estimation from populations of marked animals. Bird Study 46(Supplement):120-138.

Submitted 12 January 2011. Accepted 5 November 2013. Associate Editor was Richard T. Stevens.


Department of Environmental Science, Policy, and Management, 145 Mulford Hall, University of California, Berkeley, CA 94720

* Correspondent:

Table 1--Observations of dusky-footed woodrats Neotoms fuscipes (n =
2,864), mean number of captures per 3-day trapping session, and
number of observations per category of capture (1 = captured once,
2C = captured twice on consecutive days, 2R = captured twice with a
day of rest between captures, 3 = captured on all 3 days) by age,
sex, and reproductive status at Camp Roberts, California, 2002-

    Age-sex group           n        Mean [+ or -] SE

Adult males              1,152     1.58 [+ or -] 0.02
Adult females            1,032     1.64 [+ or -] 0.02
Reproductive females       482     1.91 [+ or -] 0.04
Juveniles                  198     1.35 [+ or -] 0.04
All woodrats combined    2,864     1.64 [+ or -] 0.01

                                    Capture category

    Age-sex group           1        2C        2R         3

Adult males                635     269        98       150
Adult females              542     246        73       171
Reproductive females       180     115        48       139
Juveniles                  140      34        12        12
All woodrats combined    1,497     664       231       472

Table 2--Mean (i.e., 10 trapping sessions combined) mass of dusky-
footed woodrats (Neotoma fuscipes) at first encounter (initial
mass), and mean change in mass per capture within trapping session
(grams and percentage) and per 3-day trapping session (grams and
percentage), by age, sex, and reproductive status at Camp Roberts,
California, 2002-2006.

                           Mean initial mass

Age-sex group             n       g [+ or -] SE

Adult males               468   227.2 [+ or -] 2.3
Adult females             417   191.2 [+ or -] 1.5
Reproductive females      262   218.7 [+ or -] 1.3
Juveniles                  46   112.2 [+ or -] 2.8
All adults combined     1,147   212.2 [+ or -] 1.2
All woodrats combined   1,193   208.3 [+ or -] 1.3

                           Mean change in mass within session (a)

                                Per capture

Age-sex group             n       g [+ or -] SE        % [+ or -] SE

Adult males               494   -7.8 [+ or -] 0.3    -3.4 [+ or -] 0.1
Adult females             487   -7.2 [+ or -] 0.2    -3.8 [+ or -] 0.1
Reproductive females      268   -5.2 [+ or -] 0.3    -2.4 [+ or -] 0.1
Juveniles                  48   -4.2 [+ or -] 0.6    -3.7 [+ or -] 0.5
All adults combined     1,249   -7.0 [+ or -] 0.2    -3.4 [+ or -] 0.1
All woodrats combined   1,297   -6.9 [+ or -] 0.2    -3.4 [+ or -] 0.1

                           Mean change in mass within session (a)

                                Per session

Age-sex group             n       g [+ or -] SE        % [+ or -] SE

Adult males               468   -10.2 [+ or -] 0.4   -4.4 [+ or -] 0.2
Adult females             417    -7.2 [+ or -] 0.2   -4.9 [+ or -] 0.2
Reproductive females      217    -7.1 [+ or -] 0.5   -3.3 [+ or -] 0.2
Juveniles                  46    -4.7 [+ or -] 1.0   -4.3 [+ or -] 0.8
All adults combined     1,102    -9.3 [+ or -] 0.2   -4.7 [+ or -] 0.1
All woodrats combined   1,148    -9.1 [+ or -] 0.2   -4.4 [+ or -] 0.1

(a) Excluding 45 observations of reproductive females who exhibited
large losses in mass (possibly due to birth) between captures.

TABLE 3--Mean percentage of change in mass and associated
standard errors for the statistically significant effects category
of capture (i.e., captured twice on consecutive days, captured twice
with a day of rest between captures, and captured on all 3 days),
season, and season x year (spring and fall 2002-2006) from the
final model for juvenile, adult male, and nonreproductive, adult
female woodrats (Neotoma fuscipes) at Camp Roberts, California,
2002-2006. Within category of capture and season, rows sharing
the same letter are not significantly different (Tukey-adjusted
multiple comparisons, [alpha] = 0.05). The significant effects plot
and plot x season are not shown.

          Variable             n         Mean [+ or -] SE

Capture category
  Twice (consecutive days)     473     -3.70 [+ or -] 0.1 a
  Twice (rest day)             162     -3.48 [+ or -] 0.3 a
  Thrice                       296     -6.80 [+ or -] 0.2 b
  Spring                       311     -4.45 [+ or -] 0.2 a
  Fall                         620     -4.75 [+ or -] 0.1 b
Season x year
  Spring 2002                    4     -6.21 [+ or -] 3.6
  Spring 2003                   36     -2.60 [+ or -] 0.5
  Spring 2004                   98     -5.57 [+ or -] 0.4
  Spring 2005                   56     -3.24 [+ or -] 0.5
  Spring 2006                  117     -4.62 [+ or -] 0.4
  Fall 2002                     64     -3.74 [+ or -] 0.5
  Fall 2003                     79     -4.76 [+ or -] 0.3
  Fall 2004                     82     -4.36 [+ or -] 0.4
  Fall 2005                    154     -5.37 [+ or -] 0.3
  Fall 2006                    241     -4.75 [+ or -] 0.2
COPYRIGHT 2013 Southwestern Association of Naturalists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Hardy, Michael A.; Zingo, James M.; Tietje, William D.
Publication:Southwestern Naturalist
Article Type:Report
Geographic Code:1USA
Date:Sep 1, 2013
Previous Article:Characterizing regeneration of woody species in areas with different land-history tenure in the Tamaulipan Thornscrub, Mexico.
Next Article:Genetic variability in natural populations of Agave durangensis (agavaceae) revealed by morphological and molecular traits.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters