population dynamics and conservation status of the ... - Oxford Journals

Journal of Mammalogy, 89(3):721–729, 2008

POPULATION DYNAMICS AND CONSERVATION STATUS OF THE INSULAR CAVY CAVIA INTERMEDIA (RODENTIA: CAVIIDAE) CARLOS H. SALVADOR

AND

FERNANDO A. S. FERNANDEZ*

Laboratorio de Ecologia e Conservaçaõ de Populaço˜es, Departamento de Ecologia, Universidade Federal do Rio de Janeiro, Rio de Janeiro 21941-590, Brazil

Cavia intermedia may be the mammal with the smallest geographical distribution in the world, because it is endemic to a 10-ha island and therefore vulnerable to extinction. The objective of this study was to understand the population dynamics of C. intermedia in order to define its conservation status and to provide information for its management. The population was sampled monthly using capture–mark–recapture methods and radiotelemetry for 16 months, between March 2004 and June 2005. A total of 5,530 trap-nights resulted in 665 captures of 141 individuals. The population exhibited the main demographic characteristics of the insular syndrome, such as high and stable density, small home ranges, stable age structure composed mostly of adults, and high survival rates. The population dynamics were characterized by density-dependent effects on mortality. C. intermedia can be considered one of the rarest species on the planet because of its limited geographic distribution and very small population size (average estimated population of 42 individuals). We recommend that it be listed as Critically Endangered. Key words: age structure, density dependence, endemism, Histricomorpha, home range, island, island syndrome, population size, rarity, wild guinea pig

The cavy Cavia intermedia Cherem, Olimpio, and Ximenez, 1999, is endemic to the 10-ha Moleques do Sul Island in southern Brazil, and may be the mammal with the smallest geographic distribution in the world, according to current studies of insular mammals (Alcover et al. 1998a, 1998b). This species has a chromosomal number different from that of any other species in its genus (Gava et al. 1998). Endemic species on islands may be particularly vulnerable (Alcover et al. 1998b; Carleton and Olson 1999; Doulan et al. 2005), thus the current population status of C. intermedia warrants investigation. Using internationally recognized criteria (World Conservation Union [IUCN] 1994, 2001), C. intermedia can be considered Vulnerable because its only known population has a very restricted area of occupancy (,2,000 ha); thus, this species could become at risk of extinction due to negative effects of human activities or stochastic events within a very short time period. Since the description of C. intermedia, no study has been carried out with the species in its natural habitat. Information concerning its biology is very scarce, especially about

its demography, and is critically needed to guide management and conservation actions. Estimating the number of individuals and assessing the probability of extinction are fundamental for an appropriate definition of the conservation status of C. intermedia (IUCN 1994, 2001), and for consideration for legal protection in Brazil. The probability of extinction can be estimated through population viability analysis, which also can provide information for managing the population (e.g., Brooks et al. 1997; Traylor-Holzer et al. 2005). However, population viability analysis requires demographic and genetic data and a precise understanding of the species’ population dynamics and threats (Miller and Lacy 2005). In principle, the population of C. intermedia should fluctuate according to the entry of individuals through recruitment and the exit of individuals through mortality (Begon and Mortimer 1986), because immigration and emigration are not probable on the island. It is supposed that C. intermedia has been restricted to the island for at least 8,000 years, since the isolation of Moleques do Sul Archipelago by lowering sea levels, which led to its speciation (Cherem et al. 1999; Gava et al. 1998). Thus, this insular population would be expected to express some particular regulatory and demographic characteristics, known as the insular syndrome (Adler and Levins 1994). The insular syndrome includes demographic characteristics such as

* Correspondent: [email protected]

Ó 2008 American Society of Mammalogists www.mammalogy.org 721

722

JOURNAL OF MAMMALOGY

Vol. 89, No. 3

FIG. 2.—Average monthly temperature and rainfall recorded between February 2004 and July 2005 on the mainland (278359S, 488349W) adjacent to Moleques do Sul Island, southern Brazil (source: Agricultural and Livestock Research and Rural Extension Company of Santa Catarina/National Institute of Meterology).

FIG. 1.—The location of Moleques do Sul Archipelago, southern Brazil.

high and stable population densities, small home ranges, high survival rates, and small variance in age structure, with the population being mostly comprised of adult individuals (Adler and Levins 1994; Gliwicz 1980). In our study, the 1st goal was to estimate values for main demographic parameters of C. intermedia in their natural habitat, such as population size, population density, survival rate, recruitment and mortality rates, age structure, and sex ratio, as well as home-range size. Based on this information, we attempted to understand the species’ population dynamics, to define its conservation status, and to provide demographic information for an eventual population viability analysis. Reproductive and growth patterns of this species are presented elsewhere (Salvador and Fernandez 2008).

MATERIALS AND METHODS Study site.—Moleques do Sul Archipelago is located 8.25 km southeast of the nearest point of the much larger Santa Catarina Island, and 14 km from the adjacent continent at Santa Catarina State, southern Brazil. The geographical coordinates are 278519S, 488269W (Fig. 1). The archipelago comprises 3 islands, formed about 8,000 years ago according to the maximum water depth between the archipelago and the continent (32 m) and estimates of sea-level variation for the region (Correâ 1996). The archipelago is part of Serra do Tabuleiro State Park. The largest island in the archipelago, Moleques do Sul, has an area of 9.86 ha, with a maximum length of 710 m and a maximum width of 200 m. It is the only site where C. intermedia is known to occur. In addition to C. intermedia, the vertebrate fauna of the island consists of 31 species of birds (Bege and Pauli 1989) plus an undescribed species of worm lizard (family Amphisbaenidae). According to field observa-

tions, the only predators of cavies on the island are raptors, including southern caracaras (Caracara plancus), yellowheaded caracaras (Milvago chimachima), and Chimango caracaras (M. chimango—Kraus and Ro¨del 2004). Two other possible predators listed in Bege and Pauli (1989), burrowing owls (Speotyto cunicularia) and peregrine falcons (Falco peregrinus), were not observed on the island during this study. With the exception of the extreme southwestern part of the island, cavies are found in all areas covered by vegetation, as indicated by the presence of their feces. These areas amount to 6.34 ha, predominantly covered by herbaceous vegetation. Cavies frequently use patches with the grasses Paspalum vaginatum and Stenotraphrum secundatum, their main food source (Cherem et al. 1999); the area of these patches amounts to 0.77 ha. Runways of cavies and a heavy concentration of their feces are found in these areas. The feeding areas are surrounded by bush and grass vegetation, especially Cortaderia selloana and Verbesina glabrata, which also are used as shelter by the cavies. The climate in the region is mesothermal humid with hotter months in the summer and rains spread over the whole year, albeit with less rainfall in the winter. During the study period, the average monthly temperature in the summer of 2005 was 258C, with a total rainfall of 585 mm for the season, according to a meteorological station located on the mainland adjacent to the island (278359S, 488349W) that belongs to the Agricultural and Livestock Research and Rural Extension Company of Santa Catarina. In the winter of 2004, the average temperature and total rainfall were 188C and 188 mm, respectively (Fig. 2). Sampling effort.— A monthly capture–mark–recapture program was carried out between March 2004 and June 2005, with 5 consecutive nights of trapping in each sampling session. The 1st trapping session in 2004 included only 4 nights, and data from this sampling period were not used in the demographic analysis. However, we included data for this month in the calculations of home-range sizes. For each trapping session, 70 wire mesh live traps (15 15 50 cm) were used, distributed around the island according to feeding areas (Fig. 3). Total trapping effort amounted to 5,530 trap-nights. The traps were arranged along runways

June 2008

SALVADOR AND FERNANDEZ—POPULATION DYNAMICS OF CAVIA INTERMEDIA

marked by the cavies (Asher et al. 2004; Kraus et al. 2003), and set at the same points in each trapping session. We assumed that all cavies on the island used the feeding areas, and therefore that the entire population was sampled by our traps. Cavies were individually marked with numbered tags (National Band and Tag Co., Newport, Kentucky). Apart from cuts in the ears (in the case of lost tags) and occasional infections, marking caused no obvious harm to the animals. To reduce stress to the animals, traps were kept shut during the day, opened by late afternoon, and examined and shut the next morning (Kraus et al. 2003), and always supplied with slices of corn. Animal care procedures were approved by the 2 environmental organizations concerned, Brazilian Institute of the Environment and Natural Renewable Resources and The State of Santa Catarina Environmental Management Foundation/ DEAM (license 101/2004 and letter 634, respectively), and met guidelines approved by the American Society of Mammalogists (Gannon et al. 2007). Population dynamics.—To estimate population parameters, we followed Pollock’s (1982) robust design, treating the population as closed when estimating the population size for each 5-day sampling session and as open when estimating recruitment and survival rates between sessions, over the 15 months from April 2004 and June 2005. We used program CAPTURE (Otis et al. 1978) to estimate population size, opting for model Mh (Burnham and Overton 1979) as the most robust for small mammals (Manning et al. 1995). Values for the minimum number known alive (MNKA—Krebs et al. 1969) were preferred when estimates by Mh were smaller than those by MNKA. We recognize that MNKA tends to produce negatively biased estimates (Nichols 1986). However, we had special interest in accurate population size estimates for conservation purposes, and when estimates by Mh were lower than the MNKA, the former must be even more negatively biased, so we opted for MNKA to minimize the negative bias. Calculations of density were based on the area of the island covered by vegetation (6.34 ha). This procedure was based on the assumption that cavies, although most often found on the feeding areas, used all the vegetated parts of the island. This assumption was reasonable because feces of C. intermedia were found in all vegetated areas and some individual cavies were captured at several different grass patches. Neither feces nor any other sign of cavies was found in the rocky parts of the island. The program JOLLY (Hines 1998) was used to estimate recruitment and survival rates through the Cormack–Jolly– Seber method. Survival rates were then standardized for a 30-day period, following the method proposed by Fernandez (1995). The standardized values were used to calculate mortality ((1 survival rate) population size). The population size was correlated with mortality and recruitment by means of Spearman correlations. The relationships between these last 2 variables and average temperature and total rainfall for each month were tested by simple linear regressions (Zar 1999). We also used linear regressions to test for lagged responses of demographic variables to rainfall, with lags from 1 to 3 months.

723

FIG. 3.—Distribution of traps () in the feeding areas of Cavia intermedia (in gray) on the largest island of Moleques do Sul Archipelago, southern Brazil, between March 2004 and June 2005.

Age categories were based on body size of cavies, where young were considered to be those weighing ,400 g, subadults those weighing between 400 and 500 g, and adults those weighing .500 g (Salvador 2006). Age structure was analyzed according to the number of animals in each class captured each month. The sex ratio (proportion of males to females) was analyzed monthly by MNKA. Binomial tests were used to analyze deviations from the expected ratio 1:1 (Zar 1999). Home range.— To estimate home-range sizes, we equipped 5 adult females and 7 adult males with radiocollars (TW-4-SM transmitters; Biotrack Co., Wareham, United Kingdom) and monitored by radiotracking (3-element yagi antenna and M-57 receiver; Biotrack Co.). Each individual was monitored every 4 h during the day. Animals were located via triangulation or visual contact. Coordinates of each location were determined in the field using a system of coordinates printed on an aerial photograph (1:1,550) of the island. Each cell in the coordinate system was 3.5 3.5 m; maximum error estimated for positions was approximately 2 m. Home-range sizes were estimated using the 100% minimum convex polygon method (Mohr 1947) in program RANGES VI (Anatrack Ltd., Warenham, United Kingdom). The monitoring interval (4 h), method, and program were based on similar studies involving other species of Cavia (Asher et al. 2004; Kraus et al. 2003). Radiotracking locations and capture points (traps) were used together in the analyses. Estimates of homerange size for each sex were compared using Mann–Whitney U-tests (Zar 1999).

RESULTS During the study, 665 captures were obtained of 141 individuals, including 79 males and 62 females (Table 1). The

724

Vol. 89, No. 3


TABLE 1.—Number of marked individuals (ni) and sex ratio for Cavia intermedia on Moleques do Sul Island, southern Brazil, from April 2004 to June 2005. Age categories (AC): young (Y), subadults (Sa), and adults (A). ni Sex

Sex ratio

AC (%)

Sex

Month (i)

#

$

2004 Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec.

10 16 9 5 12 13 16 17 17

12 4 7 8 5 8 11 6 6

3 2 1 1 1 0 0 3 3

(14%) (10%) (6%) (8%) (6%) (0%) (0%) (13%) (13%)

3 3 3 2 1 2 2 3 2

(14%) (15%) (19%) (15%) (6%) (10%) (7%) (13%) (9%)

16 15 12 10 15 19 25 17 18

2005 Jan. Feb. Mar. Apr. May Jun.

18 16 18 13 16 15

7 11 12 7 9 4

4 2 3 2 2 1

(16%) (7%) (10%) (10%) (8%) (5%)

2 1 6 2 5 3

(8%) (4%) (20%) (10%) (2%) (16%)

19 24 21 15 18 15

Total

79

62

a b c

Y

(8%)c

Sa

(13%)c

Ratio

Total

#

$

#:$

(73%) (75%) (75%) (77%) (88%) (90%) (93%) (74%) (78%)

22 20 16 13 17 21 27 23 23

10 20 21 20 22 23 21 21 25

12 10 11 12 13 16 20 14 17

1:1 2:1 2:1 2:1 2:1 1:1 1:1 3:2 1:1

(76%) (89%) (70%) (75%) (72%) (79%)

25 27 30 20a 25 19

26 25 25 19 18 15

15 15 16 12 10 4

2:1 2:1 2:1 2:1 2:1 4:1b

(79%)c

141

21c

12c

1.75:1c

A

Insufficient data to determine the age group of 1 individual. Significant difference from 1:1 in sex ratio using the binomial test (P , 0.05). Average.

sex ratio did not differ significantly from 1:1, except for during the last month when there was a bias toward males (P ¼ 0.02; Table 1). Assuming equal age-specific capture probabilities, the age structure of the population comprised (average 6 SD) of 79% 6 7% adults, 13% 6 5% subadults, and 8% 6 4% young (Table 1; Fig. 4). The average (6 SD) estimated population size was 42 6 10 individuals. Considering only the vegetated part of the island, this results in an estimated density of about 6.6 individuals/ha. The coefficient of variation of population size during the study was 24%. The average (6 SD) of the monthly survival rate estimates was 0.81 6 0.16. The averages (6 SD) for monthly recruitment and mortality rates were 10 6 8 and 9 6 10 individuals, respectively (Table 2). The population had the highest estimated mortalities, 32 and 27 individuals, in the months when population size was highest, October 2004 (62 individuals) and March 2005 (58

individuals), respectively. These 2 months were followed by population declines (Table 2; Fig. 5). Even considering time lags of up to 3 months, recruitment and population size were not correlated in any analysis (all P . 0.05). However, there was a positive correlation between population size and mortality in 2 different ways. First, population size was correlated with mortality (rs ¼ 0.60, P ¼ 0.03) that occurred 2 months previously, that is, between months i 2 and i 1. Likewise, population size also was positively correlated with mortality (rs ¼ 0.63, P ¼ 0.02) in the current month, that is, between months i and i þ 1. Recruitment, mortality, and survival were not correlated with total rainfall or temperature (all P . 0.05), even considering time lags of up to 3 months. A total of 305 locations were obtained for estimates of home-range size, resulting in an average (6 SD) home-range size of 0.17 6 0.11 ha. Each individual’s home range did not exceed 4% the size of the island. The largest home ranges recorded were 0.35 ha for male 437 and 0.34 ha for female 971. Home-range sizes did not differ significantly between sexes (U ¼ 17, P ¼ 0.94; Table 3; Fig. 6).

DISCUSSION

FIG. 4.—Age structure of the population of Cavia intermedia on Moleques do Sul Island, southern Brazil, between April 2004 and June 2005: Y ¼ young (,400 g), Sa ¼ subadults (400–500 g), and A ¼ adults (.500 g).

Comparisons of population densities among species of Cavia reported in the literature are difficult because of the diversity of methods employed for estimating population sizes and the different areas used to calculate density. For example, Rood (1972) found an average density of 20 individuals/ha in a population of C. aperea in Argentina using the number of captured individuals as an estimate of population size and the

June 2008


725

TABLE 2.—Estimates of population parameters for Cavia intermedia on Moleques do Sul Island, southern Brazil, from April 2004 to June 2005. Estimated population size (Ni) and 95% confidence interval (CI) for month i of sampling; survival rate (Øi, iþ1), recruitment (Bi, iþ1), and mortality (Di, iþ1) between month i and the next (i þ 1). Values are averages, standard deviations (SDs), and coefficients of variation (CVs). CI Month (i) 2004 Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. 2005 Jan. Feb. Mar. Apr. May Jun. X SD CV (%)

Ni

Minimum

Maximum

Øi,

a iþ1

Bi,

a iþ1

Di,

b iþ1

50 32 38 32c 35c 43 62 35 49

37 25 27 — — 32 47 28 36

76 51 62 — — 67 88 54 74

0.87 1.00 0.90 0.90 0.73 1.00 0.48 1.00 0.74

— 5 10 21 22 0 9 18 8

7 0 4 3 9 0 32 0 13

41c 40c 58 42 52 24

— — 43 31 40 21

— — 83 64 77 38

0.81 0.80 0.53 0.83 — —

7 18 2 0 — —

8 8 27 7 — —

42 10 24

— 8 26

— 15 24

0.81 0.16 20

10 8 80

9 10 111

a

Estimates made by the Cormack–Jolly–Seber method. Di, iþ1 ¼ (1 Øi, iþ1 ) Ni. c Minimum number known alive (Krebs et al. 1969); confidence intervals are not available for these estimates. The remaining estimates were carried out by Mh (Burnham and Overton 1979). b

size of the trapping grid as an estimate of sampling area. Bonaventura et al. (2003) and Asher et al. (2004) did the same for populations of C. aperea and found densities of 8 individuals/ha in Argentina and 12 individuals/ha in southeastern Brazil, respectively. Kraus et al. (2003) obtained an average density of about 14 individuals/ha for C. magna in Uruguay, using the ratio between the average MNKA and the size of the trapping grid. All these values are higher than what we considered the most reliable estimate of the average population density of C. intermedia, that is, 6.6 individuals/ha. This finding is contrary to our expectations, because 1 characteristic of the insular syndrome is high population densities (Adler 1996, 1998; Crowell 1983; Gliwicz 1984a, 1984b). On the other hand, results from the other studies may have been overestimates, as they used the size of the trapping grids— typically placed in optimum cavy habitat—for estimating densities. If we had used the same methods, considering only the size of the area trapped on Moleques do Sul (about 0.77 ha), the estimated density of C. intermedia would be 28 or 44 individuals/ha (using number of captured animals and MNKA, respectively, to represent population size). We consider these to be overestimates, because C. intermedia used other areas on the island in addition to its feeding areas. However, this comparison illustrates the point that the comparatively low

FIG. 5.—Population dynamics of Cavia intermedia on Moleques do Sul Island, Serra do Tabuleiro State Park, southern Brazil, between April 2004 and June 2005. Population size (bars), recruitment (m), and mortality (n).

density that we estimated for C. intermedia is likely to reflect methodological differences rather than a real pattern. The population of C. intermedia can be considered quite stable, with a 24% coefficient of variation in 15 months, as expected according to the insular syndrome (Adler 1996, 1998; Crowell 1983; Gliwicz 1984a, 1984b). Although Bonaventura et al. (2003) found a smaller coefficient of variation, 22% in a population of C. aperea in Argentina over 21 months of study, Rood (1972) found a much higher coefficient of variation (78%) in a population of the same species in the same country over 11 months of study, and Kraus et al. (2003) also found a higher coefficient of variation (75%) in a 21month study of C. magna in Uruguay. Temporal variations in numbers of cavies are usually associated with mortality by predation (Kraus and Ro¨del 2004) or dispersal in search of new feeding areas (Bilenca et al. 1995; Galante and Cassini 1994). Cavies may seek new areas when they suffer a shortage of resources (Bilenca et al. 1995; Bonaventura et al. 2003; Kraus et al. 2003). However, this was not the case for C. aperea in Argentina, which experienced a stable productivity throughout the year (Bonaventura et al. 2003), nor of course was it possible for for C. intermedia. TABLE 3.—Estimates of home-range size for the cavy Cavia intermedia on Moleques do Sul Island, Serra do Tabuleiro State Park, southern Brazil, between March 2004 and June 2005: individual (ID), number of locations (n), and home-range size estimates (Sˆ) in hectares (ha). ID

n

Sˆ (ha)

# 437 # 445 # 456 # 410 # 488 # 455 # 496 $ 479 $ 750 $ 971 $ 816 $ 913 X SD

46 10 31 49 28 11 31 12 18 28 15 26

0.35 0.05 0.18 0.20 0.16 0.06 0.20 0.07 0.10 0.34 0.07 0.27

25 13

0.17 0.11

726


Vol. 89, No. 3

FIG. 6.—Home ranges of different individuals of both sexes for the cavy Cavia intermedia on Moleques do Sul Island, southern Brazil, between March 2004 and June 2005. Values in meters on both axes.

For an insular population, the average monthly survival rate for C. intermedia (0.81) was high, again as expected, when compared to typical estimates on the continent. Although the population of C. magna in Uruguay had monthly survival rates of 0.94 in its most stable phase, during the 2nd half of the study the average survival rate was only 0.66 until extinction (Kraus et al. 2005), supposedly caused by predation (Kraus and Ro¨del 2004). In C. intermedia, predation should not strongly affect the population, because the main predators of cavies on the mainland (mammals) are absent. Raptors, the only potential predators on the island, are considered the least important predators of cavies on the mainland (Kraus and Ro¨del 2004).

The sex ratio remained balanced throughout the year, as noted for C. aperea in Uruguay (Kraus et al. 2003). The age structure also was as expected for a species showing the insular syndrome, with little variation and comprising mostly older individuals (e.g., Austad 1993; Gliwicz 1980, 1984b). Home-range size for C. intermedia (0.17 ha) was small in relation to home-range size for C. magna (males ¼ 1.20 ha and females ¼ 0.80 ha—Kraus et al. 2003). Home-range size for C. intermedia was more similar to that of C. aperea, which was about 0.10 ha in studies by Rood (1972; males ¼ 0.14 ha and females ¼ 0.12 ha) and Asher et al. (2004; males ¼ 0.09 ha and females ¼ 0.05 ha). A larger home-range size could have been

June 2008


expected for C. intermedia because of its phylogenetic proximity with C. magna (Cherem et al. 1999). However, smaller home ranges also are a characteristic of insular populations (Adler and Levins 1994), as a result of confinement in high densities. Thus, the population of C. intermedia manifested the main characteristics of the island syndrome (Adler and Levins 1994) regarding density (high and stable), age structure (stable and mostly composed of adults), survival rate (high), and homerange size (reduced). Salvador and Fernandez (2008) studied the reproduction and growth of C. intermedia and concluded that, even by caviomorph standards, its reproductive rates were low (78% of pregnant females with a single offspring), young are born quite large (19% of adult females’ weight), and sexual maturity is reached quite late (at around 59 days). All these patterns corroborate the idea that C. intermedia has been confined to Moleques do Sul for a long time (Cherem et al. 1999; Gava et al. 1998). On the other hand, C. intermedia must have efficient mechanisms of population regulation, according to Gliwicz (1980), to be able to sustain itself for such a long time in high and stable densities in such a small area. The slight fluctuation of the population of C. intermedia could be understood through the balance between the recruitment and mortality of individuals. The population increased with increases in recruitment in relation to mortality and decreased when the relationship between these variables was inverse (Begon and Mortimer 1986). Increased mortalities could be caused in part by per capita scarcity of resources at high population densities. According to Adler (1998), reductions in survival rate for neotropical rodents due to food resource limitations can occur not only when there is an extreme shortage (seasonal), but also when resources are abundant in growing populations. When there is either overexploitation or seasonal variations of resources, cavies usually move to areas with better conditions (Bilenca et al. 1995; Bonaventura et al. 2003; Kraus et al. 2003). On Moleques do Sul, individuals cannot disperse and would have to confront any shortage. There is some evidence of overexploitation of resources for C. intermedia, because on Moleques do Sul the grasses on which cavies mostly feed reach only about 5 cm high, whereas on other nearby islands—including the other 2 islands of Moleques do Sul Archipelago itself—they may reach 50 cm and have a higher biomass. In months after great mortality, the population of C. intermedia decreased, which led to increased survival rates, restoring the balance with recruitment. This relationship explains the positive correlation between population size and mortality. This process must be quite efficient because survival rates always responded quickly to population size. Therefore, density-dependent variations in mortality, rather than climatic effects, seem to be crucial in determining the dynamics of this population. Even though there are seasons with dry weather and low temperatures in Moleques do Sul, seasonality was not observed in the population, nor were there any correlations between climatic and population variables. Although the population of C. intermedia existed at high density during our study, the number of individuals was so small that it is probably one of the rarest of living mammal

727

species. Mammals considered exceedingly rare have dozens or even hundreds of individuals in the wild, with the rarest benefiting from protection by national laws or by international organizations such as Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) and IUCN (Massicot 2006). According to official criteria (IUCN 1994, 2001), C. intermedia should be considered in the category of most serious threat, Critically Endangered, because there are fewer than 50 adult individuals in existence. Threats to the species are related not only to its small number, but to its tiny geographic distribution as well. Because it depends on a few very small feeding areas, C. intermedia also is highly vulnerable to disturbances on the island. Based on our estimates of population size, we propose the assignment of C. intermedia to the IUCN category of Critically Endangered, and its protection by law. A population viability analysis is now feasible to further evaluate the conservation status using an additional criterion (criterion E—IUCN 1994, 2001), as well as to provide a scientific basis for practical conservation actions (e.g., Brooks et al. 1997; Traylor-Holzer et al. 2005).

RESUMO Cavia intermedia pode ser um dos mamı´feros com a menor distribuiçaõ geogra´fica do planeta por ser endeˆmica de uma ilha de 10 ha e, portanto, em elevado grau de vulnerabilidade. O estudo teve o objetivo de entender a dinaˆmica populacional para definir o estado de conservaçaõ e fornecer informaço˜es para seu manejo. A populaçaõ foi acompanhada mensalmente atrave´s de um programa de captura, marcaçaõ e recaptura e radiotelemetria por 16 meses, entre março de 2004 e junho de 2005. Um total de 5,530 armadilhas noites resultou em 665 capturas de 141 indivı´duos marcados. A populaçaõ manifestou as principais caracterı´sticas demogra´ficas da sıńdrome insular, como densidade alta e esta´vel, pequena a´rea de vida, estrutura eta´ria esta´vel e composta majoritariamente por adultos, e taxa de sobreviveˆncia alta. A dinaˆmica populacional foi melhor explicada por efeitos de denso-dependeˆncia na mortalidade. C. intermedia pode ser considerada uma das espećies mais raras do planeta, tanto pela sua limitada distribuiçaõ geogra´fica como pelo nu´mero de indivı´duos existentes na natureza (me´dia populacional estimada de 42 indivı´duos), e oficialmente deveria ser listada na condiçaõ mais grave de ameaça de extinçaõ, Criticamente Em Perigo, por crite´rios legais e reconhecidos internacionalmente.

ACKNOWLEDGMENTS We thank the financial sponsors Brazilian Coordination for Higher Level Graduates Improvement, the Research Support Foundation of Rio de Janeiro State, and Fundaçaõ O Botica´rio de Proteçaõ a` Natureza, as well as the support provided by the Department of Behavior Biology (University of Mu¨nster, Germany) in the name of Norbert Sachser and Matthias Asher, the Laboratory of Aquatic Mammals (Federal University of Santa Catarina) in name of Paulo Simo˜es-Lopes and Mauricio Graipel, Tigrinus Co. in the name of Marcos Tortato, and Caipora Cooperativa for Nature Conservation in the name of Luis Pimenta and Monica Gomes. We also thank J. D.

728


Nichols for kindly and carefully reviewing the manuscript, 2 anonymous referees for their comments and suggestions, D. Perez for donating materials, and all the colleagues that helped with field efforts.

LITERATURE CITED ADLER, G. H. 1996. The island syndrome in isolated populations of a tropical forest rodent. Oecologia 108:694–700. ADLER, G. H. 1998. Impacts of resource abundance on populations of a tropical forest rodent. Ecology 79:242–254. ADLER, G. H., AND R. LEVINS. 1994. The island syndrome in rodent populations. Quarterly Review of Biology 69:473–489. ALCOVER, J. A., X. CAMPILLO, M. MACIAS, AND A. SANS. 1998a. Mammal species of the world: additional data on insular mammals. American Museum Novitates 3248:1–29. ALCOVER, J. A., A. SANS, AND M. PALMER. 1998b. The extent of extinctions of mammals on islands. Journal of Biogeography 25: 913–918. ASHER, M., E. S. OLIVEIRA, AND N. SACHSER. 2004. Social system and spatial organization of wild guinea pigs (Cavia aperea) in a natural population. Journal of Mammalogy 85:788–796. AUSTAD, S. N. 1993. Retarded senescence in a insular population of Virginia opossums (Didelphis virginiana). Journal of Zoology (London) 229:695–708. BEGE, L. A. R., AND B. T. PAULI. 1989. As aves nas Ilhas Moleques do Sul—Santa Catarina: aspectos da ecologia, etologia e anilhamento das aves marinhas. Fundaçaõ do Meio Ambiente de Santa Catarina (FATMA), Floriano´polis, Brazil. BEGON, M., AND M. MORTIMER. 1986. Population ecology. 2nd ed. Blackwell, Oxford, United Kingdom. BILENCA, D. N., E. A. CITTADIANO, AND F. O. KRAVETZ. 1995. Influencia de la actividad de Cavia aperea sobre la estructura del habitat y la distribucioń de Akodon azarae y Oryzomys flavescens (Rodentia: Caviidae, Muridae) em bordes de cultivos de la regioń pampeana (Argentina). Iheringia 79:67–75. BONAVENTURA, S. M., V. PANCOTTO, R. L. VICARI, N. MADANES, AND M. I. BELLOCQ. 2003. Demography and microhabitat use of the wild guinea pig (Cavia aperea) in freshwater Spartina densiflora marshes in Argentina. Acta Zoologica Sinica 49:20–31. BROOKS, A. W., L. LIM, R. HARDEN, AND R. FRANKHAM. 1997. How secure is the Lord Howe Island woodhen? A population viability analysis using VORTEX. Pacific Conservation Biology 3:125–133. BURNHAM, K. P., AND W. S. OVERTON. 1979. Robust estimation of population size when capture probabilities vary among animals. Ecology 60:927–936. CARLETON, M. D., AND S. L. OLSON. 1999. Amerigo Vespucci and the rat of Fernando de Noronha: a new genus and species of Rodentia (Muridae: Sigmodontinae) from a volcanic island off Brazil’s continental shelf. American Museum Novitates 3256:1–59. CHEREM, J. J., J. OLIMPIO, AND A. XIMENEZ. 1999. Descriçaõ de uma nova espećie do geˆnero Cavia Pallas, 1766 (Mammalia—Caviidae) das Ilhas Moleques do Sul, Santa Catarina, Sul do Brasil. Biotemas 12:95–117. CORREÂ, I. C. S. 1996. Les variations du niveau de la mer durant les derniers 17.500 ans BP: l’exemple de la plate-forme continentale du Rio Grande do Sul—Bre´sil. Marine Geology 130:163–178. CROWELL, K. L. 1983. Island—insight or artifact?: population dynamics and habitat utilization in insular rodents. Oikos 41:442– 454. DOULAN, C. J., J. KNOWLTON, D. F. DOAK, AND N. BIAVASCHI. 2005. Nested communities, invasive species and Holocene extinctions:

Vol. 89, No. 3

evaluating the power of a potential conservation tool. Oecologia 145:475–485. FERNANDEZ, F. A. S. 1995. Me´todos para estimativas de paraˆmetros populacionais por captura, marcaça˜ o e recaptura. Oecologia Brasilensis 2:1–26. GALANTE, M. L., AND M. H. CASSINI. 1994. Seasonal variation of a cavy population in the Pampa region, east-central Argentina. Mammalia 58:549–556. GANNON, W. L., R. S. SIKES, AND THE ANIMAL CARE AND USE COMMITTEE OF THE AMERICAN SOCIETY OF MAMMALOGISTS. 2007. Guidelines of the American Society of Mammalogists for the use of wild mammals in research. Journal of Mammalogy 88:809– 823. GAVA, A., T. R. O. FREITAS, AND J. OLI´MPIO. 1998. A new karyotype for the genus Cavia from a southern island of Brazil (Rodentia—Caviidae). Genetics and Molecular Biology 21:77–80. GLIWICZ, J. 1980. Island populations of rodents: their organization and functioning. Biology Review 55:109–138. G LIWICZ , J. 1984a. Population dynamics of the spiny rat Proechimys semispinosus on Orchid Island (Panama). Biotropica 16:73–78. GLIWICZ, J. 1984b. Rodents on island: living in a crowd. Acta Zoologica Fennica 172:95–98. HINES, J. E. 1998. Program ‘‘JOLLY’’. United States National Biological Survey, Patuxent Wildlife Research Center, Laurel, Maryland. IUCN. 1994. IUCN Red list—categories and criteria version 2.3. IUCN Species Survival Commission, World Conservation Union (IUCN), Gland, Switzerland. IUCN. 2001. IUCN Red list—categories and criteria version 3.1. World Conservation Union (IUCN), Gland, Switzerland. KRAUS, C., J. KU¨NKELE, AND F. TRILLMICH. 2003. Spacing behavior and its implications for the mating system of a precocial small mammal: an almost asocial cavy (Cavia magna)? Animal Behaviour 66:225– 238. KRAUS, C., AND H. G. RO¨DEL. 2004. Where have all the cavies gone? Causes and consequences of predation by the minor grison for a wild cavy population. Oikos 105:489–500. KRAUS, C., D. L. THOMSON, J. KU¨NKELE, AND F. TRILLMICH. 2005. Living slow and dying young? Life-history strategy and agespecific survival rates in a precocial small mammal. Journal of Animal 74:171–180. KREBS, C. J., B. L. KELLER, AND R. H. TAMARIN. 1969. Microtus population biology: demographic changes in fluctuating populations of M. ochrogaster and M. pennsylvanicus in southern Indiana. Ecology 50:587–606. MANNING, T., W. D. EDGE, AND J. O. WOLFF. 1995. Evaluation of population-size estimators: an empirical approach. Journal of Mammalogy 76:1149–1158. MASSICOT, P. 2006. Animal info. World’s rarest mammals. www. animalinfo.org. Accessed 31 August 2006. MILLER, P. S., AND R. C. LACY. 2005. User’s manual. VORTEX: a stochastic simulation of the extinction process. Version 9.50. IUCN/SSC Conservation Breeding Specialist Group, Apple Valley, California. MOHR, C. O. 1947. Table of equivalent populations of North American small mammals. American Naturalist 37:223–249. 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.

June 2008


OTIS, D. L., K. P. BURNHAM, G. C. WHITE, AND D. R. ANDERSON. 1978. Statistical inference from capture data on closed animal populations. Wildlife Monographs 62:1–135. POLLOCK, K. 1982. A capture–recapture design robust to unequal probability of capture. Journal of Wildlife Management 46:752– 757. ROOD, J. P. 1972. Ecological and behavioral comparisons of three genera of Argentine cavies. Animal Behaviour Monographs 5: 1–83. SALVADOR, C. H. O. 2006. Biologia da conservaçaõ na teoria e na pra´tica: o estudo de caso de Cavia intermedia, um dos mamı´feros mais raros do planeta. M.S. thesis, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil.

729

SALVADOR, C. H. O., AND F. A. S. FERNANDEZ. 2008. Reproduction and growth of a rare cavy (Cavia intermedia), endemic of Moleques do Sul Island, southern Brazil. Journal of Mammalogy. TRAYLOR-HOLZER, K., R. LACY, D. REED, AND O. BYERS (EDS.). 2005. Alabama beach mouse population and habitat viability assessment: final report. IUCN/SSC Conservation Breeding Specialist Group, Apple Valley, California. ZAR, J. H. 1999. Biostatistical analysis. 4th ed. Prentice-Hall, Englewood Cliffs, New Jersey.

Submitted 10 January 2007. Accepted 30 October 2007. Associate Editor was Gerardo Ceballos.