Seasonal Survival and Cause-Specific Mortality of Urban and Rural Striped Skunks in the Absence of Rabies (2024)

Abstract

Identifying factors that are limiting to populations is fundamental to understanding population dynamics of wildlife and is important for conservation and management. Such information is of particular interest for some mesocarnivore species that are important ecologically or economically through their role as predators (Crooks and Soulé 1999; Schmidt 2003), as hosts of various diseases, and economically as furbearers or nuisances in urban and rural landscapes (Rosatte et al. 1991, 1992). Although mesocarnivores often benefit from their general food habits and habitat requirements, they also are subject to a variety of diseases, and may be affected to varying degrees by urbanization or human activities (DeStefano and DeGraaf 2003; Gehrt 2004).

Striped skunks (Mephitis mephitis) are vulnerable to a variety of mortality agents such as predation, disease, environmental conditions (e.g., severe winter or drought), chemicals, and anthropogenic activities (see Rosatte [1987] and Rosatte and Larivière [2003] for reviews). However, with the exception of rabies (Greenwood et al. 1997), the relative importance of these agents in influencing survival of skunks, and seasonal patterns of mortality, have rarely been estimated.

The striped skunk is the primary host of terrestrial rabies in the American Midwest (Parker 1962; Verts 1967; Verts and Storm 1966), and the disease typically has a profound effect on skunk abundance (Greenwood et al. 1997). This apparently was the case for skunks in Illinois, where population fluctuations are closely tied to rabies outbreaks (Verts 1967). During the 1980s and 1990s, however, statewide monitoring in Illinois failed to record a population rebound from a rabies outbreak during 1979–1980 (Illinois Department of Natural Resources, in litt.), and skunks have apparently remained at relatively low levels since then. Correspondingly, reported rabies cases have remained low to nonexistent during the same period (Illinois Department of Public Health, in litt.). The absence of rabies from the Illinois population presented the opportunity to identify alternative limiting factors for skunk populations, for which there is little information. Thus, one objective of my study was to determine seasonal and annual patterns of survival, and to identify mortality agents that may be responsible for limiting skunk populations in the absence of rabies.

For example, skunks at northern latitudes become dormant in winter dens during extreme weather (Rosatte and Larivière 2003; Verts 1967). Because they continue to metabolize energy during these dormant periods (Mutch and Aleksiuk 1977), skunks often lose weight during seasons with cold weather, which may compromise their survival. However, estimates of survival during winter, and evaluations of its effect as a potential limiting factor, are lacking for striped skunks.

A 2nd objective of my study was to determine how survival and associated mortality factors differ between skunk populations located in rural and urban landscapes. Previous studies of urban skunk populations have reported densities and movement patterns at large scales (Rosatte et al. 1991, 1992), but little is known about how mortality agents shift in importance with urbanization. The availability of artificial resources in urban systems may reduce seasonal weight loss (DeStefano and DeGraaf 2003) and improve physical condition for urban skunks during winter. Thus, survival may not be as reduced during winter for urban skunks as for rural skunks, or artificial resources may result in an overall higher annual survival for urban skunks than for rural skunks.

Alternatively, increased traffic or road systems in urban areas may increase the probability of mortality by vehicles for urban skunk populations, whereas rural skunks may be more vulnerable to hunting, trapping, or predation. In this scenario, a possible result could be similar annual survival between urban and rural populations, but shifts in predominant causes of mortality with urbanization. Herein, I compare seasonal patterns of survival and cause-specific mortality rates between skunk populations residing in rural and urban parks in Illinois.

Materials and Methods

Study areas.—I monitored skunk populations on 2 study areas 47 km apart in northeastern Illinois: the 1,499-ha Ned Brown Forest Preserve located in Cook County and the 1,214-ha Glacial Park in McHenry County. The Ned Brown Forest Preserve is located in a heavily urbanized landscape, where the predominant land use adjacent to the site is residential and industrial. Traffic volume on major roads adjacent to and bisecting the urban site ranged from 34,600 to 147,200 vehicles/24 h (Prange et al. 2003). The composition of major habitats within the study area was 39% mature woodlot, 45% open (including old-field, grassland, and mowed areas), and 16% water (including lakes and marshes). This study area had at least 32 permanent and semipermanent picnic groves scattered throughout and received intensive use by people, with more than 1 million visitors annually (Prange et al. 2003, 2004). During warm weather the picnic groves and parking lots were inundated with refuse, much of which was on the ground. The forest preserve was closed to people from sundown to sunrise throughout the year. Thus, skunks and other wildlife usually had undisturbed access to picnic groves throughout the night.

Agriculture was the predominant land use surrounding the other site, Glacial Park, primarily consisting of pasture and row crops. This rural study area was composed of 20% woodland, 70% open (primarily grassland with native grass species), and 10% water (primarily marshes). Traffic volume on major roads adjacent to the study area ranged from 900 to 5,800 vehicles/24 h (Prange et al. 2003), and human use within the park was relatively light, with only 2 picnic groves receiving minimal use (Prange et al. 2003).

The climate in northeastern Illinois is temperate, with warm summers and cool winters that can be extreme in some years. During 1999–2000, mean monthly temperatures ranged from −7.8 to 26.7°C (Illinois State Climatologist, in litt.). Cumulative snowfall ranged from 35.7 to 60.6 cm in each year, and the number of days with snowfall ranged from 18 to 20 days. Cumulative days with temperatures below 0°C ranged from 124 to 130 days each year.

Livetrapping.—Livetrapping was conducted during spring (March–April) and autumn (September–October) beginning with autumn 1998 for the rural site and spring 1999 for the urban site, and concluding in 2001 for both areas. These trapping sessions were conducted as part of a study of raccoons (Procyon lotor) described by Gehrt (2002) and Prange et al. (2003). This consisted of a 2.4-km² grid with 30 traps, with each trap (81 × 25 × 30 cm, model 108, Tomahawk Live Trap Co., Tomahawk, Wisconsin) baited with commercial canned cat food and checked each morning. In addition to this basic trapping design, concurrent supplemental trapping for skunks was conducted, where box traps (typically 12 additional traps) were opportunistically placed in areas within the trapping grid to maximize trapping success. Skunks also were captured with hand nets during spotlight sessions at night throughout each study area. Spotlight sessions usually were conducted during spring and autumn, but occasionally were conducted at other times of the year to replace radiocollars.

Captured skunks were covered by a tarpaulin (placed over the trap or net to protect from spraying) and immobilized with an injection of Telazol (Fort Dodge Animal Health, Fort Dodge, Iowa) (about 10 mg—Larivière and Messier 1996). Individuals were measured and weighed, sex was determined, and age was determined by tooth wear and reproductive condition (Verts 1967). I marked skunks with individually numbered eartags (Monel #1, National Band and Tag Company, Newport, Kentucky), and released them at their original capture sites after they had sufficiently recovered from the drug (usually within 1 h postinjection). Trapping and handling protocols followed guidelines of the American Society of Mammalogists (Animal Care and Use Committee 1998).

I partitioned body weight data into adult male, adult female, and juvenile categories during autumn, and into male and female categories during spring. I combined adult and juvenile classes in spring because few juveniles were captured during that season (1 juvenile male and 4 juvenile females), and there was little indication from this study or subsequent observations of a difference in weight between individuals of known ages. By spring, weights of juvenile skunks were at or above the springtime weights for adults of their respective sex. I compared body weights between areas and sex and age groups with analysis of variance (ANOVA) or t-tests. When significant differences were found by ANOVA, I compared means with Tukey tests.

Survival and cause-specific mortality.—I attempted to fit all captured skunks from each area with radiocollars (Advanced Telemetry Systems, Inc., Isanti, Minnesota) with motion-sensitive mortality switches (6-h delay), unless they were juveniles and weighed <2 kg during autumn (adults often weighed <2 kg in spring). Radiocollared animals were approached on foot (usually twice weekly) with a handheld antenna and receiver, each location was recorded on a map, and den sites also were noted. Radiotracking was conducted at night and skunks frequently were visually observed during these sessions. Estimated battery life for radiocollars was 9 months, and I attempted to replace collars when they neared expiration. I used fixed-wing aircraft and a helicopter, with an aerial telemetry rig, to relocate skunks if their signals disappeared from the study areas.

Animals that died were recovered as quickly as possible to facilitate necropsy and estimate survival rates. However, I generally did not try to recover skunks with mortality signals during winter denning until spring because in some instances mortality signals were emitted from dormant, but not dead, skunks. Also, recovering skunk carcasses entailed digging up the den, which might have compromised the survival of other skunks residing in the den. In these cases, deaths were attributed to disease or poor physical condition. All skunks in suitable condition were submitted to the Zoological Pathology Program at the University of Illinois Veterinary Diagnostic Laboratory for necropsies. Thorough necropsies of these carcasses determined proximate causes of mortality, such as extreme parasitism, canine distemper, rabies, and so on.

I estimated survival (S) of radiocollared skunks with the staggered entry modification to the Kaplan–Meier survival estimator (Pollock et al. 1989). Survival distributions were determined by month, and for radiocollared skunks survival was monitored throughout the month. Skunks with unknown fates were removed from the analysis during the month they disappeared. I pooled data among years to increase sample sizes, and annual survival rates were compared between areas and sex with Z-statistics, and survival distributions were compared between areas with log-rank tests (Pollock et al. 1989).

I compared cause-specific mortality rates between areas and seasons by using MICROMORT (Heisey and Fuller 1985). Because of low numbers of skunks during some months, I partitioned data into summer (April-September) and winter (October-March) seasons. Mortalities were classified as road-killed, poor physical condition or disease, predation, and unknown. As noted above, I assumed skunks recovered from winter dens had died from poor physical condition or disease. In most cases, skunks in this category (disease or poor physical condition) exhibited both characteristics. I used 95% confidence intervals (95% CIs) to compare mortality rates between seasons and populations.

Finally, as a result of the previous analyses, I used an information-theoretic approach (Burnham and Anderson 2002) to further explore possible determinants of winter survival. I used MARK (White and Burnham 1999) and a subsample of skunks to develop a series of models with covariates that might be related to winter survival. All skunks radiocollared during October, whose fates were determined from 15 October to 15 April each year, were used in the analysis. I pooled individuals across years and study areas to maximize sample size, and recorded the following covariates for each individual: sex, age, weight at autumn capture, and autumn condition index. I obtained condition indices from residuals resulting from the linear regression between body length (total length minus tail length) and weight (Green 2001) for all captured skunks during autumn trapping. I used Akaike–s information criterion adjusted for small sample sizes (AICc—Burnham and Anderson 2002) for model selection. The model with the lowest AICc is the most parsimonious model and provides the best fit to the data (Burnham and Anderson 2002). I estimated the following to assist with evaluation of models: ΔAICc gives the difference in AICc between the current model and the best model (lowest AICc), and Akaike weights (w) assess the relative support that a given model has from the data, compared to the other models in the set (Burnham and Anderson 2002).

Results

I captured and radiocollared 73 skunks, of which 36 were captured at the urban site and 37 at the rural site. Mean weight of urban males was greater (t = −2.74, d.f. = 22, P = 0.011) than that of urban females during spring (Fig. 1). During autumn, urban adult males and females were heavier than juveniles of both sexes (F = 9.3, d.f. = 2, 25, P < 0.001), but no difference (Tukey test, P > 0.05) was found in mean weight between the sexes. At the rural site, weights of males were similar to weights of females during spring (t = 1.7, d.f. = 19, P = 0.11), but during autumn adult males were heavier than adult females and juveniles of both sexes (F = 8.8, d.f. = 2, 20, P = 0.002).

Skunks of both populations exhibited weight loss during winter, resulting in lower weights in spring than in autumn (Fig. 1). At the urban site, mean weights of females declined 28% from autumn to spring (t = −2.7, d.f. = 28, P = 0.011), whereas mean weights of males declined 19% (t = −2.1, d.f. = 12, P = 0.063). Weights of rural females declined 32% over winter (t = 5.8, d.f. = 12, P < 0.001), whereas weight of rural males declined 41% (t = 5.2, d.f. = 17, P < 0.001).

I recorded 30 confirmed mortalities; additionally, 7 individuals slipped out of their collars and were not recaptured, and I lost the signals of 8 skunks (Table 1). Of the skunks that were lost, most were probably due to transmitter expiration because I frequently recaptured skunks with expired transmitters (duration of battery life in the transmitters was 9 months). Others may have dispersed, although I used a fixed-wing aircraft or helicopter multiple times to locate dispersers and rarely located skunks outside study areas.

Skunk survival distributions were similar (χ² = 0.10, d.f. = 1, P > 0.1) between areas. Both populations exhibited marked declines in survival during winter and early spring (Fig. 2). Little difference was found in annual survival (S) between areas (rural, S = 0.42, 95% CI = 0.29–0.60; urban, S = 0.40, 95% CI = 0.27–0.59; Z = 0.57, P = 0.57). After pooling data between areas, annual survival of males (S = 0.41, 95% CI = 0.24–0.58) was less (Z = −2.23, P = 0.013) than that of females (S = 0.51, 95% CI = 0.36–0.65). Both sexes had similar (χ² = 1.18, d.f. = 1, P = 0.55) seasonal patterns of survival.

For urban skunks, cause-specific mortality rates were similar (χ² = 0.80, d.f. = 3, P = 0.85) between seasonal intervals, with disease or poor physical condition the most common recorded source of mortality in both intervals (Table 2). However, the distribution of mortality rates differed slightly (χ² = 7.20, d.f. = 3, P = 0.06) between seasons for rural skunks. At the rural site, mortality related to disease or poor physical condition was higher (Z = 2.56, P = 0.01) during winter and spring than during summer and autumn. Comparisons of mortality rates between areas yielded similar estimates during winter and spring (P > 0.1 for all), but disease or poor physical condition was greater (Z = 2.55, P = 0.01) for urban skunks than for rural skunks during summer and autumn.

Brain tissue from 12 skunks submitted for necropsy tested negative for rabies by the fluorescent antibody examination, indicating that they were not infected with the virus. Pneumonia was observed in 6 of 16 necropsies, even for those individuals whose proximate cause of death was vehicles (3 of 6) or predators (1 of 1). The most common type of pneumonia was verminous pneumonia, although it was not always possible to determine the cause of pneumonia (histopathology was often inconclusive because of autolysis, which had commonly occurred in skunks recovered from winter dens). Necropsies also revealed skunks from both areas had infections of multiple endoparasites (>8 species; not all parasites were identified to species), with the most common being Physaloptera (62%), Skrjabingylus chitwoodorum (56%), Dipetalonema (31%), Baylisascaris columnaris (25%), Capillaria (25%), and Sarcocystis (25%). Each skunk was infected with at least 1–6 different species of endoparasites. One-half of the skunks had a parasitic infection of such severity that it probably contributed to their demise (as diagnosed by the pathologist), and 4 (50%) of these were collected between January and March. Lice (Trichodectes) were prevalent among the skunks in this study, but did not appear to compromise survival.

I recorded known fates during winter and early spring seasons for 35 skunks for which I also had covariate data. Model selection did not clearly indicate any model as a best fit to the data. The model with only sex as a covariate had the lowest AICc value (Table 3), but its Akaike weight (w) was only 0.18, identical to that of the null model S (.). ΔAICc values also suggested that there was little difference between models in fit, which ranged between 0 and 0.02 for the next-best models (Table 3).

Discussion

Seasonal survival patterns in this study suggested that mortality factors were greater in winter in both study populations, despite the differences in anthropogenic effects and general ecology between study areas. Skunks do not hibernate but remain dormant in winter dens (referred to as winter denning) for extended periods dictated by weather (Rosatte and Larivière 2003; Verts 1967). Winter and early spring seasons during this study appeared to represent a challenge to skunk survival during winter denning. The weight loss I recorded in this study, presumably a result of extensive dormant periods, was similar to that reported in previous studies. Average weight loss was 49% and 58% during 2 winters for skunks that had extensive denning periods in Minnesota (Sunquist 1974), and winter weight losses were 38% for females and 14% for males in New York (Hamilton 1937). When using data from captive animals, Mutch and Aleksiuk (1977) estimated that skunks experience an energy deficit of 44% below that needed to maintain body mass if they maintain a basal metabolic rate for an extended winter denning period of 140 days.

Despite considerable attention to the winter denning ecology of skunks by naturalists in the 20th century (Allen and Shapton 1942; Dean 1965; Jones 1939; Selko 1938; Sunquist 1974), few estimates of winter survival are available, especially when using radiotelemetry. In Saskatchewan, Larivière and Messier (1998) reported approximately 40% mortality during summer, which is considerably higher than our estimates for the same season. In North Dakota, spring–summer survival of skunks was 0.85 before a rabies outbreak, and only 0.17 during the outbreak the following year (Greenwood et al. 1997). When using carcass data and assuming a stable population, Fuller and Kuehn (1985) estimated an annual survival rate of 0.67 for adults and 0.17 for juveniles in Minnesota. Their estimate for juveniles includes possible mortality in utero and in the natal den; their overwinter survival rate for juveniles was 0.29.

The combination of relatively low survival and significant weight loss suggests that skunks undergo stress during winter. Other studies have provided indirect or anecdotal evidence as to the stresses of winter on skunks. Sunquist (1974) reported that 6 of 22 radiotagged skunks died during extensive (averages of 90–115 consecutive days) winter denning periods in Minnesota. Of 12 skunks recovered from excavated winter dens in Michigan, 3 were dead (1 possibly from a vehicle). One of the dead skunks was in a den occupied by another skunk, and it was partially eaten (Allen and Shapton 1942), as was one of the skunks in Sunquist's (1974) study.

Despite considerable differences in urbanization, human activity, and traffic volumes between study areas, patterns of weight fluctuations, survival distributions, and cause-specific mortality were similar. Likewise, population densities varied little between areas. Seasonal capture–mark–recapture estimates for seasons with suitable recapture rates ranged between 2.1 and 5.9 skunks/km² for the urban population and 2.5 and 5.5 skunks/km² for the rural population, and estimates of minimum population sizes were nearly identical between populations for every season (Gehrt 2004).

Body weights and seasonal weight fluctuations were similar between populations, unlike those of raccoons reported for the same areas (Prange et al. 2003). Although skunks probably consumed trash opportunistically, they were only occasionally observed exploiting trash at picnic groves in the urban park during tracking sessions. The lack of reliance on refuse by skunks was in stark contrast to raccoons that resided in the study area. Raccoons dramatically altered their foraging patterns in response to refuse (Prange et al. 2004), which resulted in more dramatic weight losses in winter for raccoons in the urban area relative to raccoons in the rural study area (Prange et al. 2003). This reflected the seasonal nature of human visitation to the urban site, with few visits during winter months and, consequently, less refuse. Nevertheless, raccoons continued to forage in and around garbage cans during winter despite the sharp reduction in garbage, which exacerbated their weight loss relative to that of rural raccoons (Prange et al. 2003, 2004).

Like raccoons (Prange et al. 2003), the difference in traffic volume between study areas did not result in an increase in road mortality for urban skunks. Many (n = 16) skunks in the urban area did not cross roads with high levels of vehicular traffic, so that roads formed the boundaries of their home ranges (Gehrt 2004). Apparently the traffic volume or road width (roads ranged from 4 to 10 lanes) was such that it deterred some skunks from crossing. A similar number (n = 16) from the rural sample also did not cross roads; however, there were large areas outside the park that skunks could traverse and not encounter roads. Similar rates of vehicular mortality between populations appear to be a response by skunks to recognize and avoid roads with heavy traffic volumes.

Lower survival by males relative to females also may be a function of winter stress. During winter, skunks den communally in winter dens, presumably for thermal conservation (Mutch and Aleksiuk 1977; Verts 1967). In my study, I observed the composition of these communal groups often contained multiple females, but usually no more than 1 male (see also Rosatte and Larivière 2003). Males apparently are intolerant of other males at winter dens, with the result that some males must den solitarily during winter (unless the population sex ratio is highly skewed toward females).

It is notable that predation was a rare event in both populations. Coyotes (Canis latrans) were present on both study areas throughout the study period. Indeed, at least 2 packs were resident in the urban site throughout this study, and prewhelping density was conservatively estimated to be 1.1 coyotes/km² in 2000 (Gehrt 2004), and the population exceeded 2 coyotes/km² in subsequent years. Thus, opportunities for intraguild competition certainly existed during this study. Some studies have suggested that coyotes limit populations of mesopredators such as striped skunks (Crooks and Soulé 1999; Rogers and Caro 1998), but I found no evidence that coyotes depredated large numbers of skunks on either study area during the study period. Coyote predation also was a minor cause of mortality for raccoons in the same study areas (Prange et al. 2003). The relationship between coyotes and other mesopredators, particularly those outside the canid axis, needs rigorous evaluation (Gehrt and Clark 2003).

Ironically, communal denning during winter, which might be an adaptive behavioral response (Mutch and Aleksiuk 1977), also may facilitate the spread of infectious diseases (Houseknecht 1969). Rosatte et al. (1986) suggested that communal denning may be responsible for endemic rabies for skunks on the northern prairies, and Greenwood et al. (1997) stated that communal denning also may be responsible for rabies outbreaks in the same system. Consequently, communal denning may represent an evolutionary trade-off for skunks, with the benefit of thermal conservation in the den mitigated by the costs of disease transmission.

Although rabies was not present during my study, skunks in both populations apparently were exposed to, and infected by, other pathogens potentially compromising skunk survival. A large number of skunks were suffering from pneumonia and severe infections from parasites. The parasites we identified in this study have been previously reported in skunk populations (Stegeman 1939; Verts 1967), but the relative degree of infestation at the individual level, and the possible limiting role of parasites at the population level, have only been assessed for rabies. In addition to necropsies, I observed that serology from these populations revealed high seropositive rates for canine distemper (>80%), canine parvovirus (>80%), and some degree of exposure to variants of leptospirosis (2–17% among 6 serovars).

Interactions between winter weather, den suitability, physical condition, and infection from a variety of pathogens may explain why I failed to find covariates such as condition indices as important for winter survival. I observed differential survival among individuals within denning groups, which suggests that infectious disease may be less of an issue than the nutritional plane and parasite burden of an individual at the onset of winter denning. However, it is possible that certain diseases with low virulence during most of the year may challenge an individual in declining condition during winter, particularly if immunities concurrently decline. Although rabies was absent during this study, other disease-causing agents may still play a role in population dynamics, and may explain why I failed to find a relationship between autumn body condition and winter survival. Results of necropsies suggested that most skunks were infected with some form of pneumonia and multiple macroparasites at all times of the year. Some parasite burdens were quite heavy, even during late winter or early spring as body condition of skunks deteriorated. Given the plethora of pathogens involved in the ecology of skunks, multiple proximate factors may be at play each year, but the ultimate cause of mortality is a declining nutritional plane as a result of winter dormancy. A variety of pathogens, rather than a single disease-causing agent, may be a prominent cause of mortality in skunk populations in the absence of rabies.

Acknowledgments

I am grateful to the Illinois Department of Natural Resources-Furbearer Fund, the Max McGraw Wildlife Foundation, and the Forest Preserve District of Cook County for support. The Max McGraw Wildlife Foundation, the Forest Preserve District of Cook County, and McHenry County Conservation District allowed access to their properties. R. Bluett provided the initial motivation, and subsequent advice and support throughout, and C. Anchor (Forest Preserve District of Cook County) and his staff facilitated and assisted with multiple aspects of this project. Necropsies were performed by M. Kinsel and colleagues with the Zoological Pathology Program, and M. Behr and colleagues, all associated with the Veterinary Diagnostic Laboratory, University of Illinois. The manuscript benefited considerably from the constructive comments of 2 reviewers. For brevity it was necessary to for me to refer to myself with regard to fieldwork; however, many individuals assisted with fieldwork, including T. Dolbeare, T. Preuss, K. Van Why, M. Cramers, G. Turschek, and especially D. Bogan. I appreciate their willingness to sacrifice clothes and other items due to the notable defensive mechanism exhibited by skunks.

Literature Cited

Author notes

Associate Editor was Douglas A. Kelt.

© 2005 American Society of Mammalogists