Habitat use by Oncifelis guigna and Pseudalopexculpaeus in a fragmented forest landscape in centralChile

GERARDO ACOSTA-JAMETT1,2,* and JAVIER A. SIMONETTI1

1Departamento de Ciencias Ecologicas, Facultad de Ciencias, Universidad de Chile, Las Palmeras

3425, Nunoa, Casilla 653, Santiago, Chile; 2Current address: U.G. Patrimonio Silvestre, Corporacion

Nacional Forestal, Region del Biobıo, Casilla 5, Chillan, Chile.; *Author for correspondence (e-mail:

gacosta@conaf.cl)

Received 22 August 2002; accepted in revised form 22 April 2003

Key words: Carnivores, Conservation, Forest fragmentation, Habitat selection, Oncifelis guigna,

Pseudalopex culpaeus, Scent-stations

Abstract.. Native forests in Chile have been replaced by exotic plantations, generating a mosaic of

forest fragments immersed in a matrix of plantations and farms. This landscape change could affect

habitat availability for native carnivores such as Oncifelis guigna (Felidae) and Pseudalopex culpaeus

(Canidae). We determine the potential impact of native forest fragmentation on these species, assessing

habitat use by these carnivores. The study was carried out in continuous forest (150 ha), native forest

fragments (< 10 ha) and surrounding pine plantations in central Chile (358590 S, 728410 W). We placed

scent-stations (SSs) in three types of habitat from January to August 2000, setting a total of 620 SSs to

determine habitat selection. We characterized structure and composition of 100 m2 of vegetation around

every SS, and measured its distance to nearest native forest patch, road and bottom creek. O. guigna

preferred habitats with dense bush cover, far from roads and close to large patches of native forest, being

almost exclusively restricted to this type of habitat. On the other hand, P. culpaeus preferred open

habitats, close to roads, relying extensively on pine plantations but scarcely using forest and forest

fragments. Our results suggest that native forest fragmentation, considering habitat use only, would have

a negative effect on O. guigna populations. Thus, forest size reduction and isolation may reduce O.

guigna survival in a pine dominated landscape. In contrast, P. culpaeus, a generalist species that would

use habitat depending on its abundance, may adapt to changes in this landscape.

Introduction

Habitat loss has widespread consequences for vertebrates (Caughley 1994; Fahrig

and Merriam 1994). Animal vulnerability to deforestation and forest fragmentation

depends on habitat preference and requirements. Forest specialists will be more

affected by fragmentation, due to the reduction of available area in which to

maintain a minimum viable population (Caughley 1994; Songer et al. 1997). On the

other hand, generalist species, which use the original forest as well as the new

anthropogenic matrix of surrounding habitat, will be less influenced (Virgos 2002;

Virgos et al. 2002). Specialist mammalian carnivorous species could be particularly

affected due to their need for extensive areas (Beier 1993; Maehr and Cox 1995;

Lidicker and Koenig 1996).

# 2004 Kluwer Academic Publishers. Printed in the Netherlands.

Biodiversity and Conservation 1135–1151, 2004.13:

Deforestation and forest fragmentation is a growing phenomenon in the Neotropics

(Skole and Tucker 1993), with a 0.4% annual rate (FAO 1999). Forests in Chile are not

an exception (Donoso and Lara 1996; Lara et al. 1996). In south-central Chile (368 to

388 S, 728 W), native forests have been extensively and intensely fragmented, and only

a few tracts remain in creeks and protected areas surrounded by a matrix of exotic pine

plantations [Pinus radiata (San Martın and Donoso 1996; Bustamante and Castor

1998; Grez, Bustamante, Simonetti, Fahrig 1998]. In the coastal area of central Chile,

31% of the native forest was replaced by P. radiata in only 9 years (Lara et al. 1996).

This transformation implies a threat to vertebrate survival. In fact, this area has the

greatest conservation problems in Chile (Simonetti 1994, 1999). All mammalian

carnivorous species living in this area, i.e. Pseudalopex culpaeus, P. griseus, Oncifelis

guigna, G. cuja, P. concolor, O. colocola, and Conepatus chinga, are of conservation

concern (Glade 1988). Of these threatened carnivores, O. guigna is a small felid

occurring in south-central Chile and in adjacent areas of Argentina (Cabrera 1960;

Redford and Eisenberg 1992). O. guigna is currently considered to be endangered in

Chile (Glade 1988) because of hunting and habitat destruction (Miller et al. 1983) and

to be vulnerable in Argentina (Dıaz and Ojeda 2000) but indeterminate by IUCN

(Nowell and Jackson 1996). The species is a native forest specialist (Greer 1966;

Dunstone et al. 2002a; Sanderson et al. 2002), inhabiting southern temperate forests

and semi-open areas (Miller, Rottman, Raedeke, Taber 1983). In fact, in fragmented

areas of Chiloe Island, southern Chile, O. guigna utilizes corridors of native vegetation

to move between forest fragments (Sunquist and Sanderson 1998; Sanderson et al.

2002). Thus, and considering that most small felids that live in forests, like O. guigna,

usually hunt small types of prey above the trees and need understory vegetation for

cover while stalking prey and breeding (Kleiman and Eisenberg 1973; Palomares et al.

2000), we may infer that this felid would be negatively affected by progressive de-

struction of native forests and replacement by plantations, threatening its persistence in

the long term due to the scarcity of available habitat to maintain viable populations.

Pine plantations might not be a suitable habitat for O. guigna given the scarcity of

understory vegetation because of silvicultural activities (c.f., Palomares et al. 2000).

On the other hand, P. culpaeus is a generalist canid, thriving in different habitats

such as steppe, scrub and native forest (Medel and Jaksic 1988; Jaksic, Jimenez

et al. 1990; Martınez et al. 1993; Jimenez et al. 1996; Novaro 1997). Furthermore,

P. culpaeus even inhabits pine plantations (Munoz and Murua 1990). Therefore, it

should be less affected by forest fragmentation. Nevertheless, given the paucity of

studies in central Chile, its conservation status remains unknown (Glade 1988).

Vegetation and landscape variables could influence habitat use of both species. In

this regard, structural features of habitats, such as cover, diversity and density of

tree species, and variables at landscape level, such as presence of roads and distance

to native forest, might influence animal preference or avoidance in a given habitat.

Moreover, in the coastal central region of Chile, plantations have arisen in

mountainous areas where creeks often harbor remnants of native vegetation, which

could act as a corridor for forest specialist carnivores. If O. guigna is a forest

specialist, we would expect it to be more abundant in large tracts of native forest

than in pine plantations, because of the heterogeneity of the former. In contrast,

1136

P. culpaeus should use all landscape patches and use forest in proportion to habitat

abundance.

Here we test the hypothesis that carnivorous species inhabiting fragmented

forests in a Pinus-dominated landscape are influenced by fragmentation. Specifi-

cally, we determine whether both species are restricted to some vegetation type and

elaborate on future conservation status.

Methods

Study site

Fieldwork was conducted in the coastal area of the Maule region of central Chile

(358590 S, 728410 W), at Los Queules National Reserve and in the exotic pine

plantations and fragments of native forest surrounding it (Figure 1). This region was

originally covered by Nothofagus glauca forests (Gajardo 1994). However, over the

last 25 years P. radiata and Eucalyptus globulus plantations have replaced native

forest. Currently, pines cover almost 80% of the region (CONAF-CONAMA 1998).

Our study site covered an area of 680 ha. Of these, 330 ha were pine plantations,

230 ha were continuous native forest (Los Queules National Reserve and private

forests adjoining it) and 120 ha were native fragments scattered in a matrix of pine

plantations, as follows (Figure 2):

(a) Continuous forest: native forests located at Los Queules National Reserve

(150 ha; CONAF 1999) and privately owned native forests form the largest

track of continuous native forest available (over 600 ha), covering almost 34%

of the study site. Vegetation is dominated by N. glauca, Cyptocarya alba and

Aetoxicon punctatum.

(b) Forest fragments: this habitat covers 18% of the study site (120 ha). This habitat

corresponds to native forest remnants of different size, ranging from 1 to 10 ha.

Vegetation is dominated by N. glauca, N. obliqua and C. alba.

(c) Pine plantations: this habitat covers 48% of the study area (330 ha). Vegetation

is P. radiata with occasional understory Teline montpessulana, Rosa moschata

and Aristotelia chilensis.

Habitat use

Habitat use by O. guigna and P. culpaeus was assessed through the scent-station

(SS) method (Linhart and Knowlton 1975; Roughton and Sweeny 1982; Conner

et al. 1983). We determined the use of continuous forest, forest fragments and pine

plantations by the number of stations with animal tracks. Each station was made by

a sifted soil 1 m diameter where in its center we placed a small pastille of gypsum

(25 mm diameter and 5 mm height), saturated with Bobcat Urine as scent attractants

(Leg Up Enterprise: 10 drops/tablet) (Linhart and Knowlton 1975). This technique

has been used in various studies to determine relative abundance of carnivores

1137

(Lindzey et al. 1977; Roughton and Sweeny 1982; Conner et al. 1983; Travaini et al.

1996), including Chilean carnivores (Jimenez et al. 1991; Rau et al. 1991; Munoz-

Pedreros et al. 1995; Jimenez et al. 1996), and for attracting both felid and canid

species (Jimenez et al. 1991; Rau et al. 1991).

A total of 32 SS were installed in accordance with the relative abundance of the

different habitat types across the landscape (Figure 2). Eleven stations were set in a

continuous forest, 6 on forest fragments and 15 on pine plantations. Stations were

placed 0.3 km apart, a distance regarded adequate to consider stations as in-

dependent replicates in forested environments (Nottingham et al. 1989; Jaksic et al.

1990; Jimenez et al. 1991, 1996). SSs were operated from 3 to 5 nights. They were

set in the afternoon and checked the next morning (Linhart and Knowlton 1975). A

total of 620 SS/nights were active from January to July 2000, in seven sample

Figure 1. Study area showing Los Queules National Reserve where the study was conducted. Major

towns (squares), rivers and Highway 5 are shown.

1138

periods: 47% (286) were in pine, 35% (219) in forest and 18% (115) in fragments.

Tracks were identified to species level following Acosta and Simonetti (1999).

Vegetation analysis

To determine if O. guigna and P. culpaeus are indeed associated with a particular

kind of vegetation structure and composition, we determined five vegetation

variables within 100 m2 from each SS: trees- and shrub-species richness, stem

density and vertical foliage cover: less than 1.5 m, from 1.5 to 6 m and more than

6 m in height (see Jimenez et al. 1991).

Figure 2. Study area; positions of SSs are indicated by black dots. Fifteen SSs were set up in pine

plantations, 11 on forest reserve and 6 on forest fragments. The line delimits Los Queules National

Reserve (CONAF 1999).

1139

Landscape features

Each SS was geo-referenced with a GPS and located on a map (scale 1:20 000). We

evaluated the (a) distance from each SS to the nearest road, (b) distance from each

SS to the center of Los Queules National Reserve, and (c) distance from each SS to

the nearest creek, after Estades and Temple (1999).

Habitat selection

Habitat selection was estimated following a Compositional Analysis, in order to

avoid potentially non-independent data. Habitat availability was assessed as the

proportion of active SS per each habitat type in each sample period over the total

number of SS in that period. Utilized habitat was assessed as the proportion of SSs

visited (Aebischer et al. 1993). We transformed the available and utilized habitat

composition to log-ratios, employing the formula di¼ ln(Xui/Xuj)� ln(Xai/Xaj),

where Xui is the proportion of a species’ habitat use in a habitat i and Xai is the

proportion of the available habitat i in the sample period. To rank the habitat type in

order of use, a matrix was created. Each habitat type was matched with the other two

habitat types considered in the sample period. When di> 0 it implies that relative to

habitat j, habitat i is used more than expected. Equivalently, the use of habitat j

relative to habitat i is less than expected. Conversely, di< 0 for all would imply that

habitat j was the relatively most used habitat type (Aebischer et al. 1993). The

number of positive elements in each row was assessed to rank the habitats in order of

increasing relative use, where 0 is ‘less used’ and 2 is ‘more used’.

According to Aebischer et al. (1993), we replaced a 0% of visit to a given habitat

type in a period for a non-zero value, to allow log-ratio transformations. For O.

guigna and for P. culpaeus, 0% values were replaced by 0.01 and 0.001, respec-

tively. These figures are an order of magnitude lower than the smallest non-zero

record in a given habitat (0.12 and 0.06%, respectively).

Subsequently, we performed a multiple regression to predict the probability of a

carnivore visit to an SS as a function of habitat and landscape attributes (Zar 1984).

For this purpose, a proportion of visit to each SS in a sample period was regarded as

a dependent variable, and vegetation and landscape factors as independent vari-

ables. We test the effect of each variable and perform a forward stepwise regression

using those variables not correlated themselves.

Results

Habitat description

Vegetations in continuous and forest fragments have more tree species richness and

greater stem density than those in pine plantations (Tukey’s HSD, P< 0.05; Table

1). Cover differed among habitats (F4, 87¼ 7.4; P� 0.001). Pine plantations had

1140

significantly less cover at <1.5 and 1.5–6.0 m layers (Tukey’s HSD, P� 0.001).

Continuous forest and forest fragments had similar cover in all strata, except at the

upper level (Tukey’s HSD, P< 0.02; Figure 3).

Landscape attributes differed among pine plantations, continuous forest and

forest fragments (F4, 87¼ 13.4; P� 0.001). SSs in pine plantations and fragments

were two times farther away from bottom creeks than continuous forest (Tukey’s

HSD, P< 0.001). SSs placed in pine plantations and native forest fragments were

equally distant from Los Queules National Reserve (Tukey’s HSD, P< 0.001).

Finally, SSs placed in pine plantations were closer to roads than those placed in

continuous forest and forest fragments (Tukey’s HSD, P� 0.001; Figure 4).

Habitat use and composition

We recorded a total of 58 visits of carnivores, of which 44.8% (26) were O. guigna,

41.4% (24) P. culpaeus, 12.1% (7) skunk, C. chinga, and 1.7% (1) a mustelid, G.

cuja in a total of 620 SSs/nights (Table 2). It seems that visits by one species were

not affected by another because several times one SS was visited by two different

species the same night.

The overall comparison of habitat use with habitat availability in the study area

for O. guigna and P. culpaeus differed significantly from random (w22¼ 40.5,

P< 0.0001 and w22¼ 25.3, P< 0.001, respectively). The matrix constructed ranked

O. guigna habitat in the order: continuous forest � fragment > pine plantations

(Tables 3 and 4). On the other hand, the matrix ranked P. culpaeus habitat as: pine

plantation � fragment > continuous forest (Table 5).

The univariate linear regression model showed that the relative abundance of O.

guigna was correlated with the proximity of Los Queules National Reserve. Fur-

thermore, O. guigna showed a significant relationship with high-density vegetation

cover in all vertical layers. This indicates that O. guigna abundance is more likely

in areas that are close to a big native forest patch as with high vertical cover,

preferentially in the understory (Table 6).

The multivariate analysist conducted with variables that influenced visits to SS,

showed that the most important variable to explain the abundance of O. guigna was

a high vertical cover at less than 1.5 m. (Table 7).

Table 1. Characterization of woody species found in all three habitat types.

Variables Habitat type F2, 59 P

Pine Continuous Forest

plantation forest fragment

mean� 1 SE mean� 1 SE mean� 1 SE

Richness of woody species 1.3� 0.2 6.3� 0.6 7.7� 0.8 51.1 �0.001

Density (tree/100 m2) 6.1� 0.9 25.4� 3.0 27.7� 5.9 21.1 <0.001

1141

Figure 3. Mean percentage of vegetation cover in three kinds of habitat in three vertical layers.

Different letters show statistical differences between habitat types inside each layer (�1 SE).

1142

Figure 4. Mean distance of SSs placed on three different habitat types (pine, forest and fragments) to

landscape attributes considered. (A) Distance to nearest bottom creek, (B) distance to Los Queules

National Reserve, and (C) distance to closest road. Different letters show statistical differences between

habitat types (�1 SE).

1143

On the other hand, relative abundance of P. culpaeus was strongly related to

places relatively open in all vertical layers, with few species and stem density

(Table 8). The multivariate analysis conducted showed that abundance of P.

culpaeus depends mainly on places with less cover between 1.5 and 6.0 m

(Table 9).

Table 2. Carnivore visits to SSs placed in three habitat types. Figures are

the number of visits and the percentage of each species in a given habitat

(in parentheses).

Species Visit (%) Total

Forest Fragments Pine

O. guigna 20 (64.5) 4 (80) 2 (9) 26 (44.8)

P. culpaeus 5 (16.1) 1 (20) 18 (81.8) 24 (41.4)

C. chinga 5 (16.1) 0 2 (9) 7 (12.1)

G. cuja 1 (3.2) 0 0 1 (1.7)

Total 31 5 22 58

Table 3. Ranking matrices for O. guigna: (a) matrix constructed for the

first sample period in Appendix 1(a); (b) matrix of means and standard

errors obtained by averaging each element over all seven periods.

Habitat type

Pine Forest Fragment

(a) Period 1

Pine �0.294 �0.876

Forest 0.294 �0.582

Fragment 0.876 0.582

(b) Means� 1 SE

Pine �3.04� 0.82 �1.69� 0.54

Forest 3.04� 0.82 1.36� 0.74

Fragment 1.69� 0.54 �1.36� 0.74

Table 4. Simplified ranking matrices for O. guigna comparing propor-

tional habitat use with proportion of total available habitat types. Each

mean element in the matrix was replaced by its sign; a double sign

represents significant deviation from random at P< 0.05.

Habitat type Habitat type Rank

Pine Forest Fragment

Pine �� �� 0

Forest þþ þþ 2

Fragment þþ �� 1

1144

Discussion

Habitat selection depends on the scales at which an organism samples its en-

vironments (Krebs et al. 1974). At small scales, animal physiological and per-

ceptual abilities are involved, rendering habitat selection an organism-specific

feature (Kotliar and Wiens 1990). Several habitat features constrain an individual’s

Table 5. Simplified ranking matrices for P. culpaeus comparing

proportional habitat use with proportion of total available habitat types.

Each mean element in the matrix was replaced by its sign; a double sign

represents significant deviation from random at P< 0.05.

Habitat type Habitat type Rank

Pine Forest Fragment

Pine þþ þþ 2

Forest �� � 0

Fragment �� þ 1

Table 6. Effects of each vegetation and landscape variable on O. guigna

index of abundance in coastal central Chile, as assessed from the results of

a linear regression univariate model. Asterisks indicate variables retained

for multivariate logistic regression analysis. b¼ regression coefficient.

SE(b)¼ standard error of beta.

Variables Coefficient (b) SE(b) P

Vegetation

Percentage of cover <1.5 m 0.23 0.07 �0.01*

Percentage of cover from 1.5 to 6 m 0.16 0.07 <0.05*

Percentage of cover >6.0 m 0.22 0.07 <0.01*

Stem density 0.13 0.07 0.06

Tree species richness 0.11 0.07 0.11

Landscape

Distance to nearest creek �0.12 0.07 0.07

Distance to Los Queules NR �0.25 0.06 �0.01*

Distance to closest road �0.12 0.07 0.07

Table 7. Best multivariate regression model describing relationship

between habitat and landscape condition and SS O. guigna index of

abundance in coastal central Chile (F3, 220¼ 7.37; P� 0.001). b¼ logistic

regression coefficient. SE(b)¼ standard error of beta.

Related variables Coefficient (b) SE(b) P

Percentage of cover <1.5 m 0.14 0.08 0.04*

Distance to LQ NR 0.10 0.07 0.21

Percentage of cover >6.0 m �0.15 0.08 0.06

1145

decisions regarding its use, such as food availability, predation risk, the presence of

competitors, etc. (e.g., Neu et al. 1974). Dunstone et al. (2002a) have suggested

that, in southern Chile, O. guigna prefers habitats with high cover and high prey

abundance, which can represent recognition of a patch structure at a low hierarchy

level (Johnson 1980). On the other hand, Jimenez et al. (1996) pointed out that P.

culpaeus exploits habitat resources at small scales, preferring habitat with slopes.

Our results strongly suggest that O. guigna prefers large tracks of native forests,

then forest fragments and, lastly, pine plantations, as depicted for other carnivores

like badgers, martens and wildcats (Virgos et al. 2002). This differential use might

account for vegetation characteristics rather than prey availability. O. guigna preys

on rodents and birds, where their main preys are small arboreal and terrestrial

mammals like Akodon olivaceus, Irenomys tarsalis and Dromiciops australis

(Dunstone et al. 2002b). In our study area, both rodent and bird abundance is higher

in forest fragments and plantations than in continuous forest (Saavedra and Si-

monetti submitted; Vergara 2002), as often found in other temperate fragmented

landscapes (Tellerıa et al. 1991; Tew and Macdonald 1993). Therefore, prey

availability might not be the main factor triggering the reliance of O. guigna on the

Table 8. Effects of each vegetation and landscape variable on P. culpaeus

index of abundance in coastal central Chile, as assessed from the results of

a linear regression univariate model. Asterisks indicate variables retained

for multivariate logistic regression analysis. b¼ regression coefficient.

SE(b)¼ standard error of beta.

Variables Coefficient (b) SE(b) P

Vegetation

Percentage of cover <1.5 m �0.15 0.07 0.02*

Percentage of cover from 1.5 to 6 m �0.22 0.07 �0.01*

Percentage of cover >6.0 m �0.19 0.66 <0.01*

Stem density �0.14 0.07 0.04*

Tree species richness �0.14 0.07 0.03*

Landscape

Distance to nearest creek 0.03 0.07 0.71

Distance to Los Queules NR �0.04 0.07 0.55

Distance to closest road �0.08 0.07 0.21

Table 9. Best multivariate regression model describing relationship

between habitat and landscape condition and SS P. culpaeus index of

abundance in coastal central Chile (F3, 220¼ 5.18; P< 0.001). b¼ logistic

regression coefficient. SE(b)¼ standard error of beta.

Related variables Coefficient (b) SE(b) P

Percentage of cover 1.5� 6.0 m �0.34 0.14 <0.01*

Tree species richness 0.19 0.13 0.15

Percentage of cover >6.0 m �0.09 0.07 0.20

1146

continuous forest. Habitat with a high cover in the understory might provide O.

guigna protection from hunters and a suitable breeding habitat, as is the case for the

Iberian lynx, Lynx pardinus, in Mediterranean scrublands (Rodrıguez and Delibes

1992; Palomares et al. 2000). The reduced presence of O. guigna in plantations may

also be explained by the intense human presence with high road density for ser-

vicing plantations, but also offering easy access to humans and their related ac-

tivities (i.e. dogs, illegal hunting), rendering plantations an unsafe habitat.

On the other hand, foxes are generalist species able to use a wide range of habitat

(Chadwick et al. 1997). In this sense, P. culpaeus would be employing different

types of habitat depending on its availability. In our pine plantation-dominated

landscape, foxes are more commonly found in plantations, because this type of

habitat is the most abundant, but also, as pointed out by Munoz and Murua (1990),

due to the abundance of roads, which improve movement capabilities and allow

them to move through plantations to hunt small mammals.

Habitat fragmentation resulting from forestry and clear-cutting may well result in

range expansion of generalist species (Songer et al. 1997). In our study site, P.

culpaeus preys on rodents present on pine plantations (i.e. Octodon bridgesi). Thus,

P. culpaeus could benefit from the increase in abundance of prey associated with

pine plantations and fragments (Saavedra and Simonetti submitted). Therefore,

habitat fragmentation and replacement by itself might not have a negative impact

on P. culpaeus. On the other hand, the conservation status of O. guigna would

depend on distribution and abundance of large forest patches remaining in the area.

Therefore, if we consider O. guigna as a native forest specialist, its conservation

status should be critical due to fragmentation, which has been extensive in central

Chile (Donoso and Lara 1996), reducing native forest cover and therefore potential

O. guigna habitat. Nevertheless, pine plantations placed around native forest are not

an abrupt boundary (Palomares et al. 2000), as they are utilized by O. guigna, but it

depends on how close they are to forest (Estades and Temple 1999). Thus, frag-

ments that remain surrounded by a matrix of plantations could permit movements

between native forest patches (Fahrig and Merriam 1985; Stamp et al. 1987).

Habitat selection in carnivorous species sheds light on habitat use and move-

ments in a new landscape, which could be utilized by managers and forestry

companies to render these habitats compatible with conservation strategies of some

native species like O. guigna, maintaining forest fragments across the large areas

covered by pine plantations in order to allow its movement across a low quality

habitat such as pine plantations.

Acknowledgements

This study was part of a Master Thesis of G.A. at the Faculty of Science, University

of Chile. The research was funded by Fondecyt 1981050 (J.S.) and Beca PG/47/

2000 University of Chile (G.A.). Forestal Millalemu and Conaf kindly granted

permission to conduct research in their properties. P. Vergara, R. Bustamante and J.

Armesto and two anonymous reviewers made helpful comments on earlier drafts.

1147

References

Acosta G. and Simonetti J.A. 1999. Guıa de huellas de once especies de mamıferos del bosque templado

chileno. Boletın del Museo Nacional de Historia Natural (Chile) 48: 19–27.

Aebischer N.J., Robertson P.A. and Kenward R.E. 1993. Compositional analysis of habitat use from

animal radio-tracking data. Ecology 74: 1313–1325.

Beier P. 1993. Determining minimum habitat areas and habitat corridors for cougars. Conservation

Biology 7: 94–108.

Bustamante R.O. and Castor C. 1998. The decline of an endangered temperate ecosystem: the ruil

(Nothofagus alessandrii) forest in central Chile. Biodiversity and Conservation 7: 1607–1626.

Cabrera A. 1960. Mamıferos sudamericanos: Historia Natural. Editorial Ediar, Buenos Aires, Argentina

Caughley G. 1994. Directions in conservation biology. Journal of Animal Ecology 63: 215–244.

Chadwick A.H., Hodge S.J. and Ratcliffe P.R. 1997. Foxes and forestry. Technical Paper No. 23. Forestry

Commission, Edinburgh, UK.

CONAF 1999. Plan de manejo Reserva Nacional Los Queules. Documento de Trabajo No. 314. Cor-

poracion Nacional Forestal, Talca, Chile.

CONAF-CONAMA 1998. Catastro de Bosque nativo en la VIII region. Corporacion Nacional Forestal,

Concepcion, Chile.

Conner M.C., Labisky R.F. and Progulske J.R. 1983. Scent-station indices as measures of popula-

tion abundance for bobcats, racoons, gray foxes and opossums. Wildlife Society Bulletin 11:

146–152.

Dıaz G.B. and Ojeda R.A. (eds) 2000. Libro rojo de mamıferos amenazados de la Argentina. SAREM,

Buenos Aires, Argentina.

Donoso C. and Lara A. 1996. Utilizacion de los Bosques Nativos de Chile: pasado, presente y futuro. In:

Armesto J.J., Villagran C. and Arroyo M.K. (eds) Ecologıa de los bosques nativos de Chile, Editorial

Universitaria, Santiago, Chile, pp. 367–387.

Appendix 1. Percentage of habitat composition within each habitat type for (a) O. guigna and (b) P.

culpaeus.

Sample period number Habitat type

Pine Forest Fragment

(a) O. guigna data

1 33.33 33.33 33.33

2 0.00 100.00 0.00

3 0.00 87.50 12.50

4 0.00 75.00 25.00

5 0.00 100.00 0.00

6 16.67 66.67 16.67

7 0.00 0.00 0.00

Habitat composition (% of study area) 47 34 19

(b) P. culpaeus data

1 100.00 0.00 0.00

2 0.00 0.00 0.00

3 68.75 25.00 6.25

4 0.00 0.00 0.00

5 85.71 0.00 14.29

6 100.00 0.00 0.00

7 50.00 50.00 0.00

Habitat composition (% of study area) 47 34 19

1148

Dunstone N., Durbin L., Wyllie I., Freer R., Acosta G., Mazzolli M. and Rose S. 2002a. Spatial

organization, ranging behaviour and habitat utilization of the Oncifelis guigna in southern Chile.

Journal of Zoology (London) 257: 1–11.

Dunstone N., Freer R., Acosta-Jamett G., Durbin L., Wyllie I., Mazzolli M. and Scott D. 2002b. Uso del

habitat, actividad y dieta de la guina (Oncifelis guigna) en el Parque Nacional Laguna San Rafael, XI

Region, Chile. Boletın del Museo Nacional de Historia Natural (Chile) 51: 147–158.

Estades C.F. and Temple S.A. 1999. Deciduous-forest bird communities in fragmented landscape

dominated by exotic pine plantations. Ecological Applications 9: 573–585.

Fahrig L. and Merriam G. 1985. Habitat patch connectivity and population survival. Ecology 66:

1762–1768.

Fahrig L. and Merriam G. 1994. Conservation of fragmented populations. Conservation Biology 8:

50–59.

FAO 1999. States of the World’s Forest. FAO, Rome, Italy.

Gajardo R. 1994. La vegetacion natural de Chile: clasificacion y distribucion geografica. Editorial

Universitaria, Santiago, Chile.

Glade A. (ed) 1988. Libro rojo de los vertebrados terrestres de Chile. Corporacion Nacional Forestal,

Santiago, Chile.

Greer J.K. 1966. Mammals of Malleco Province, Chile.Publication of the Museum, Michigan State

University, Biological Series 3: 51–151.

Grez A.A., Bustamante R.O., Simonetti J.A. and Fahrig L. 1998. Landscape ecology, deforestation, and

forest fragmentation: the case of the ruil forest in Chile. In: Salinas-Chavez E. and Middleton J. (eds)

Landscape Ecology as a Tool for Sustainable Development in Latin America. http://www.brocku.ca/

epi/lebk/grez.hmtl.

Jaksic F.M., Jimenez J.E., Medel R.G. and Marquet P.A. 1990. Habitat and diet of Darwin’s fox

(Pseudalopex fulvipes) on the Chilean mainland. Journal of Mammalogy 71: 246–248.

Jimenez J.E., Marquet P.A., Medel R.G. and Jaksic F.M. 1991. Comparative ecology of Darwin’s fox

(Pseudalopex fulvipes) in mainland and island setting of southern Chile. Revista Chilena de Historia

Natural 63: 177–186.

Jimenez J.E., Yanez J.L., Tabilo E.L. and Jaksic F.M. 1996. Niche-complementarity of South American

foxes: reanalysis and test of an hypothesis. Revista Chilena de Historia Natural 69: 113–123.

Johnson D.H. 1980. The comparison of usage and availability measurement for evaluating resource

preference. Ecology 61: 65–71.

Kleiman D.G. and Eisenberg J.F. 1973. Comparisons of canid and felid social systems from an evolu-

tionary perspective. Animal Behaviour 21: 637–659.

Kotliar N.B. and Wiens J.A. 1990. Multiple scales of patchiness and patch structure: a hierarchical

framework for the study of heterogeneity. Oikos 59: 253–260.

Krebs J.R., Ryan J.C. and Charnov E.L. 1974. Hunting by expectation or optimal foraging? A study of

patch use by chickadees. Animal Behaviour 22: 953–964.

Lara A., Donoso C. and Aravena J.C. 1996. La conservacion del bosque nativo en Chile: problemas y

desafıos. In: Armesto J.J., Villagran C. and Arroyo M.K. (eds) Ecologıa de los Bosques Nativos de

Chile. Editorial Universitaria, Santiago, Chile, pp. 335–362.

Lidicker Jr. W.Z. and Koenig W.D. 1996. Response of terrestrial vertebrates to habitat edges and

corridors. In: McCullough D.R. (ed) Metapopulation and Wildlife Conservation. Island Press, Wa-

shington, DC. pp. 85–109.

Lindzey F.G., Thompson S.K. and Hodges J.I. 1977. Scent-station and it use for estimation of black bear

abundance. Journal of Wildlife Management 41: 151–153.

Linhart S.B. and Knowlton F.F. 1975. Determining the relative abundance of coyotes by scent-station

lines. Wildlife Society Bulletin 3: 119–124.

Maehr D.S. and Cox J.A. 1995. Landscape features and panthers in Florida. Conservation Biology 9:

1008–1019.

Martınez D.R., Rau J.R. and Jaksic F.M. 1993. Respuesta numerica y selectividad dietaria de zorros

(Pseudalopex spp.) ante una reduccion de sus presas en el norte de Chile. Revista Chilena de Historia

Natural 66: 195–202.

1149

Medel R.G. and Jaksic F.M. 1988. Ecologıa de los canidos sudamericanos: una revision. Revista Chilena

de Historia Natural 61: 67–79.

Miller S.D., Rottmann J., Raedeke K.J. and Taber R.D. 1983. Endangered mammals of Chile: status and

conservation. Biological Conservation 25: 335–352.

Munoz A. and Murua R. 1990. Control of small mammals in a pine plantation (central Chile) by

modification of the habitat of predators (Tyto alba, Strigiformes and Pseudalopex spp., Canidae). Acta

Oecologica 11: 251–261.

Munoz-Pedreros A., Rau J.R., Valdebenito M., Quintana V. and Martınez D.R. 1995. Densidad relativa

de pumas (Felis concolor) en un ecosistema forestal del sur de Chile. Revista Chilena de Historia

Natural 68: 501–507.

Neu C.W., Byers R.C. and Peek J.M. 1974. A technique for analysis of utilization–availability data.

Journal of Wildlife Management 38: 541–545.

Nottingham Jr. B.G., Johnson K.G. and Pelton M.R. 1989. Evaluation of scent-station surveys to monitor

racoon density. Wildlife Society Bulletin 17: 29–35.

Novaro A. 1997. Pseudalopex culpaeus. Mammalian Species 558: 1–8.

Nowell K. and Jackson P. 1996. Wild Cats. Status Survey and Conservation Plan. IUCN, Gland,

Switzerland.

Palma L., Beja P. and Rodrıguez M. 1999. The use of sighting data to analyse Iberian lynx habitat and

distribution. Journal of Applied Ecology 36: 812–824.

Palomares F., Delibes M., Ferreras P., Fedriani J.M., Calzada J. and Revilla E. 2000. Iberian lynx in a

fragmented landscape: predispersal, dispersal, and postdispersal habitats. Conservation Biology 14:

809–818.

Rau J., Martınez D.R., Wolfe M.L., Munoz-Pedreros A., Alea J.A., Tillerıa M.S. and Reyes C.S. 1991.

Predacion de pumas (Felis concolor) sobre pudues (Pudu pudu): rol de las liebres (Lepus europaeus)

como presas alternativas. In: Oltremari J. (ed) Proceedings of II International Congress of Manage-

ment in Natural Resource, Valdivia, Chile, pp. 311–331.

Redford K.H. and Eisenberg J.F. 1992. Mammals of the Neotropic. Vol. 2: The Southern Cone.

University of Chicago Press, Chicago, Illinois.

Rodrıguez A. and Delibes M. 1992. Current range and status of the Iberian lynx Felis pardinus

Temminck, 1824 in Spain. Biological Conservation 61: 189–196.

Roughton R.D. and Sweeny M.W. 1982. Refinements in scent station methodology for assessing trend in

carnivore populations. Journal of Wildlife Management 46: 217–229.

Sanderson J., Sunquist M.E. and Iriarte A. 2002. Natural history and landscape-use of guignas (Oncifelis

guigna) on Isla Grande de Chiloe, Chile. Journal of Mammalogy 83: 608–613.

San Martın J. and Donoso C. 1996. Estructura florıstica e impacto antropico en el bosque Maulino de

Chile. In: Armesto J.J., Villagran C. and Arroyo M.K. (eds) Ecologıa de los Bosques Nativos de Chile.

Editorial Universitaria, Santiago, Chile, pp. 153–168.

Simonetti J.A. 1994. Threatened biodiversity as an environmental problem in Chile. Revista Chilena de

Historia Natural 67: 315–319.

Simonetti J.A. 1999. Diversity and conservation of terrestrial vertebrates in Mediterranean Chile. Revista

Chilena de Historia Natural 72: 493–500.

Skole D. and Tucker C. 1993. Tropical deforestation and habitat fragmentation in the Amazon: satellite

data from 1978 to 1988. Science 260: 1905–1910.

Songer M.A., Lomolino M.V. and Perault D.R. 1997. Niche dynamics of deer mice in a fragmented, old-

growth-forest landscape. Journal of Mammalogy 78: 1027–1039.

Stamp J.A., Buechner M. and Krishnan V.V. 1987. The effects of edge permeability and habitat geometry

on emigration from patches of habitat. American Naturalist 129: 533–552.

Sunquist M. and Sanderson J. 1998. Ecology and behaviour of the O. guigna in a highly-fragmented,

human dominated landscape. Cat News 28: 17–18.

Tellerıa J.L., Santos T. and Alcantara M. 1991. Abundance and food searching intensity of wood mice

(Apodemus sylvaticus) in fragmented forests. Journal of Mammalogy 72: 183–187.

Tew T.E. and Macdonald D.W. 1993. The effects of harvest on arable wood mice, Apodemus sylvaticus.

Biological Conservation 65: 279–283.

1150

Travaini A., Laffitte R. and Delibes M. 1996. Determining the relative abundance of European red foxes

by scent-station methodology. Wildlife Society Bulletin 24: 500–504.

Virgos E. 2002. Are habitat generalists affected by forest fragmentation? A test with Eurasian badgers

(Meles meles) in coarse-grained fragmented landscapes of central Spain. Journal of Zoology (London)

258: 313–318.

Virgos E., Tellerıa J.L. and Santos T. 2002. A comparison on the response to forest fragmentation by

medium-sized Iberian carnivores in central Spain. Biodiversity and Conservation 11: 1063–1079.

Wiens J.A. 1973. Pattern and process in grassland bird communities. Ecological Monographs 43:

237–270.

Zar J.H. 1984. Biostatistical Analysis. 2nd edn. Prentice-Hall, Englewood Cliffs, New Jersey.

1151