Rates and causes of mortality in a fragmented population of Iberian lynx Felis pardina Temminck, 1824

Дата канвертавання19.04.2016
Памер181.26 Kb.

Rates and causes of mortality in a fragmented population of Iberian lynx Felis pardina Temminck, 1824

P. Ferreras, J. J. Aldama, J. F. Beltrán & M. Delibes

Estación Biologica de Doñana, CSIC, Apdo. 1056, 41080 Sevilla, Spain

Data are presented on mortality rates of the Iberian lynx Felis pardina in Doñana

National Park and the surrounding area in southwestern Spain, collected from

1983 to 1989. Data from radio-tagged lynxes and records of lynx mortality were used to assess causes of death. Annual mortality rate, estimated by using the

Heisey and Fuller methodology, was 037. Human activities were the main cause of lynx mortality in this mainly fully protected area, either directly (illegal trap­

ping: 417%; road traffic: l67%; hunting with dogs: 83%) or indirectly (artesian

wells: 8.3%). Only 83% of the annual mortality can be related unequivocally to natural causes. Sex, age, social status, year, and home range situation influenced mortality rates. The trend of this lynx population during the seven years of study revealed serious threats to its survival. This case study exemplified how a small population can approach extinction because of increased mortality due to habitat fragmentation, despite full protection of the main portion of its range.

Many authors have referred to the extinction of species caused by habitat fragmentation (Lovejoy et at., 1984; Diamond 1989). However, there are few descriptions of the extinction process of these fragmented populations. In this paper we analyse the mortality of an isolated and fragmented popu­ lation of Iberian lynx Felis pardina in risk of ex­ tinction (Palomares et aL, 1991).

The Iberian lynx is considered the most endan­ gered carnivore in Europe (Mallinson, 1978). The total population has been estimated to be about

1000 individuals, distributed in many small and isolated populations in the southwestern Iberian Peninsula (Rodriguez & Delibes, 1992). The Dofiana lynx population was estimated at 40—50 individuals in 1987 (Palomares et a!., 1991), its range including the protected Dofiana National Park (550 km2) and the recently created Dofiana Environment Natural Park (540 km2).

Nevertheless, this lynx population has recently

decreased in numbers (Rau et at., 1985; Palomares et a!., 1991). Hence, it is important to know the rates and causes of mortality, in order to attempt to reduce them.

The objectives of this study were fourfold:

(1) to estimate annual mortality rate of the

Doflana lynx population from 1983 to 1989; (2) to estimate differential mortality rates by sex, age, social status, year and home range


(3) to determine causes of mortality; and

(4) to discuss the ability of the current Doflana protected area to enhance lynx survival.

Estimation of mortality rate and anecdotal infor­ mation on birth rates can allow us to speculate about the annual balance and trends of the Doflana lynx population, and assess management plans for the conservation of the Iberian lynx.

Located in southwestern Spain, the study area covers approximately 2750 km2, including Doflana


Fig. 1. Sketch of the study area. Road C-445 is marked with an asterisk and the tourist urbanization with a double circle.

National Park (550 km2) and the recently created (1989) Natural Park of the Dofiana Environment (540 km2) (see Palomares et at 1991). There are many detailed descriptions of this zone (Amat et at, 1979; Rogers & Myers, 1980). The Doflana area, flat and low, is on the western bank of the Guadalquivir river (Fig. 1). The area is also bordered on the south and west by the Atlantic Ocean, on the north by the alluvial plain of the Tinto river and the intensively cultivated high ground of the Aljarafe, and on the east by the Marismas of the Guadalquivir river.

The land bordering the protected area is under considerable human exploitation. A large develop­ ment of coastal tourism (in the Matalascaflas urbanization), intensively cultivated (olive trees, vines) and irrigated (fruit trees and strawberries) lands, pine Pinus pinea plantations exploited for wood and edible pine seeds in the west and euca­ lyptus plantations in the north, have increased human pressure in the area in the last 25 years. Several roads connecting villages, especially the

C-445 from El Rocio to Matalascaflas, cross habi­

tats normally occupied by lynxes.

Three lynx nuclei are distinguished by Palo-

mares et at (1991) in the population under study: a large central one occurring mainly in the National Park, but effectively divided into two subnuclei by areas of open marsh and cultivation; a second in the northwest (Moguer), and the smallest one in the east (Puebla). The last two subpopulations are mainly in the unprotected area.

Between 1983 and 1989 we radio-tagged 30 lynxes, representing 9523 transmitter-days (Delibes & Beltrán, 1986). Animals were captured with trunk- traps, box-traps and victor No. 2 coil-spring traps (Woodstream Co., USA) with cushioned jaws to avoid injuries. They were immobilized with a mix­ ture of Ketamine and Xylazine, weighed, mea­ sured, equipped with radios, and samples of blood

N deams Moflality fate

[ N deaths H Mo’tatty rate


measurements (Beltrán, 1988). By using radio- tracking data on the lynx space-use patterns, we have distinguished three social classes in the Dofiana lynx population: residents (individuals holding a permanent home range); dispersers (in­ dividuals with no established home range, fre­ quently juveniles and subadults dispersing from their natal ranges, but also some adults ejected from their home ranges by competitors); and pre­ dispersant young (kittens and juveniles that live in their natal range) (Griffith & Fendley, 1986; De­ libes, 1989). Mortality rates were estimated for each class of age, sex and social status.

Fig. 2. Variation of the Annual Mortality Rate throughout the study period and number of total dead lynx (with and without transmitter) found each year. The number of radio- tracking days are given in parentheses.

and exoparasites taken. Mortality rates were esti­

mated using the MICROMORT program (Heisey

& Fuller, 1985), based on calculating daily mortal­

ity rates, according to the total number of radio­

transmitter days and deaths of radio-tracked ani­ mals occurring in the interval covered (Trent & Rongstad, 1974). The different rates were com­

pared at the p < 001 level with the test proposed by these authors in the program manual. This em­ ploys a statistic based on the corresponding rate estimates, their variances and the covariance (all calculated by the program), and fits into a Z-normal distribution.

All records of dead lynxes in the Dofiana area in the period 1983—89 were used to estimate mor­

tality causes and their relative importance.

Individual lynxes were separated into three age

classes: juvenile (under one year old), subadult U—2 years old), and adult (over two years old), ac­ cording to dentition, weight and other physical

Annual mortality rate
Average annual mortality rate (AMR) from radio- tracking data was 037 ± 0025 (mean ± standard error) with high variability during the study period (SD = 026; Fig. 2), and significant differences be­ tween 1984 and 1986 (Z = 320, p = 0001) and

1986 and 1989 (Z 239, p = 0008). Mortality seemed to be cyclical, with peaks in 1984 and 1988

and a minimum in 1986. Social status appeared to have the strongest effect on mortality rate (Table 1),

dispersers having the highest mortality (male AMR = 086, female AMR = 070). As stated, dispersers included adults expelled from their territory (two cases, found dead in 1984 and 1987). Predisperser individuals had the lowest mortality rate, although in this group only individuals older than five months are included, since animals younger than this age cannot be fitted with a transmitter. None of the seven young male lynxes radio- tracked died in their natal range, before dispers­

Table 1. Annual mortality rates (AMR), 95% confidence intervals (CI), number of radio-tracking days (RD), number of tracked indi­

viduals (IND) and number of recorded deaths (DEA) by different categories of sex and social status

Sex and social status


95% Confidence





Resident male






Resident female






Predispersing male






Predispersing female






I)ispersing male






I)ispersing female






Total male






Total female





The total number of males and females does not coincide with the addition of numbers in each category because some individuals were tracked while belonging to different categories.

Table 2. Causes of mortality: annual mortality rates (AMR) due to each cause, percentage (%) of the total AMR, confidence inter­ vals (CI), number of deaths for the radio-tagged sample (DEAR); total of deaths (DEAT) (tagged and not tagged) and percentage (%) of total deaths due to each cause





CI 95%




Illegal trapping


















































ing, and only one female out of the eight radio- tagged individuals died (Table 1). Mortality rates were greater in males than in females, though not

significantly (Z = O9l,p = 0184).
Causes of death
During the seven years of study we recorded 24 cases of dead lynxes (Table 2). The main cause of mortality in this period was accidental illegal trap­ ping in coil-spring traps and snares set to catch foxes and rabbits. At least eight individuals were killed in this way, of which four had been radio- tagged (AMR = 015). A female lynx with kittens was killed by a poacher with hunting dogs. The second most important cause of mortality was road traffic (AMR = 006). Five lynxes were killed by cars on road C-445, which runs along the west­ ern limit of the National Park, crossing lynx habi­ tats. At least three lynxes drowned in artesian wells (12 5%, the first of these radio-tagged; after dis­ covering it, we looked in other wells in the area and found two more). Four individuals were found dead from unknown causes (167%). Another lynx was caught alive by men and probably died of stress (the ‘accident’ in Table 2). Two adult females died of starvation due to natural causes. One was radio- collared and was lame through arthrosis, while the other had two broken ribs probably due to being kicked by a red deer Cervus elaphus. Evidence of prey-caused mortality in wolves Canis lupus is given in Mech and Nelson (1990).

The AMR of the Doflana lynx population is high, and similar to the mortality of some harvested

bobcat Felis rufus populations (e.g. Fuller et a!.,

1985). This suggests that being categorized as an endangered species (IUCN, 1988), protected by Spanish law, and occurring in protected areas, does not guarantee the conservation of this feud.

Most lynx mortality in Doñana is caused by ac­

cidents related to human activities, with about

75% of the AMR either directly (illegal trapping, road accidents, hunting with dogs) or indirectly (drowning in artesian wells) man-caused. This is a consequence of the high humanization of the area surrounding the National Park (a similar case is described by Mech (1989) for the wolf). Though illegal, the use of coil traps and snares to kill rab­ bits and foxes in many private properties border­ ing the National Park (mostly agricultural land) is common. Road traffic is also very high, especially in summer, because of intensive coastal tourism.

Inter-annual variations in mortality rates could be

caused by changes in rabbit availability, as rabbits are the staple prey of Iberian lynx (Delibes, 1980), and rabbit breeding depends on weather conditions (Soriguer, 1981; Vandewalle, 1989). Scarcity of rab­ bits must favour lynx dispersal and then increase mortality rates. However, at present we are not able to demonstrate a cause—effect relationship be­ tween rabbit numbers and lynx mortality.

Fragmentation increases the number and length

of edges (Laurence & Yensen, 1991) and is associ­ ated with increased mortality risk to lynxes. The risk of living in the surrounding area is shown when the mortality rate of adult resident males (0.48, n = 566 radio-transmitter days) in The Acebuche (a protected property by the western side of the Park bordered by the C-445 road) is compared with the mortality rate of adult males living in the Biological Reserve, in the middle of the park (0•00, n = 2225 radio-transmitter days).

Fragmentation of good-quality habitats also ex­ plains the high mortality of dispersers, which are forced to cross high-risk habitats when looking for new areas to settle. It is interesting to note that females, which have shorter and more de­ layed dispersal movements than males (authors, unpublished data), have lower mortality rates than males, although this difference is not significant.

As noted, we have not radio-tagged individuals under five months of age. Hence, we do not know mortality rates of kittens under this age, although for some similar species, such as Felis lynx, this time is very critical (Kaczensky, 1990). We have only two records of dead lynx under five months old: one died of unknown causes, while the other was illegally trapped. Also field behaviour (activ­ ity movement patterns) of some radio-tagged lac­ tating females seems to indicate high neonatal mortality (authors, unpublished data). On the contrary, mortality in predispersing individuals more than five months old is very low.

Based on radio-tracking data and anecdotal records of reproduction, we can speculate about birth rates and evaluate the extreme values of the annual balance of the Doñana lynx population, assuming it to be isolated (Palomares et aL, 1991; Rodriguez & Delibes, 1992, this issue). We assume a mean population size of 45 individuals (40—50, according to the cited authors), with 10—13 repro­ ductive females. Female productivity will depend on food availability and other factors (diseases, predators, hazards), and will never exceed two young reaching five months of age per female per year. Using our estimated mean annual mortality rate from the radio-tagged sample (AMR037), the net annual increase will rarely reach nine indi­ viduals per year (the difference between the maxi­ mum total recruitment, 13 X 2 = 26 five month- old kittens, and the estimated number of deaths,

45 X 037 = 167). However, in unfavourable years for breeding (when few or no females are able to breed or kitten mortality is very high), the population could lose 17 lynxes (the mean number of annual deaths), a third of its total number. These simple calculations suggest a pessimistic fu­ ture for the Doñana lynx population, because the number of individuals lost in unfavourable years exceeds the potential increase in optimal ones.

Total protection of the lynx range in Doflana does not ensure the survival of the population. Some years ago, Dasmann (1983) stated that

‘areas designated for nature conservation . . . sur­

rounded by lands used intensively . . . may be in-

adequate to provide for the survival of the species originally contained within them’. Our study cor­ roborates this point of view and provides quanti­ tative information to analyse the relationship be­ tween fragmentation, size and shape of a protected area, and mortality of a top predator. At Doñana, lynx are not being hunted toward extermination as happened in the past with Felis lynx in Central Europe (Eiberle, 1972). Rather, they are dying be­ cause of diverse human activities which are not directed against them (illegal trapping focused on foxes and rabbits). Most deaths occur during dis­ persal, when lynx cross the limits of the protected area. Conservation of biodiversity in the Doñana National Park, as illustrated in this case study of the Iberian lynx, requires a policy of nature and wildlife conservation throughout the region, not only inside the protected area, which is too small to hold a minimum viable population of the species (Shaffer & Samson, 1985; Palomares et al.,

1991). The recent creation of the Natural Park of

the Doñana Surroundings is a necessary, but still insufficient, step forward.


This study has been supported by the CAICYT­ CSIC (project No. 944), FONDENA-WWF Spain and the DGICYT-CSIC (project No. PB87—0405). We thank R. Laffitte, who has performed almost all the captures and tracked the lynxes. We are also grateful to N. Bustamante for English correction. Dr P. Jackson, Dr E. Heligren and an anonymous referee have kindly commented on the manuscript.
Amat, J. A., Montes del Olmo, C.. RamIrez Djaz, L., Torres Ramirez, A. (1979). Parque Nacional de Doñana. Mapa ecologico. ICONA, Madrid.

Beltrán, J. F. (1988). EcologIa y conducta espacio-temporal del lince ibérico (Lynx pardina Temminck, 1824) en el Parque Nacional de Doñana. PhD thesis, University of Seville.

Dasmann, R. F. (1983). The relationship between protected areas and indigenous peoples. In Proceedings of the World Congress on National Parks. ed. J. A. McNeely & K. R. Miller. IUCN, Gland, Switzerland, pp. 667—71.

Delibes, M. (1980). Feeding ecology of the Spanish lynx in the Coto Dofiana. Acta Theriologica, 25(24), 309—24.

Delibes, M. (1989). Factors regulating a natural population

of Iberian lynxes. Conf Reintroductions of Predators in

Protected Areas, Torino, Italy, 1987, pp. 96—9.

Delibes, M., Beltrãn, J. F. (1986). Radio-tracking of six species of carnivores in the Donana National Park, SW Spain. Mesogée, 46, 113—20.

Diamond, J. (1989). Overview of recent extinctions. In Con­

servation for the Twenty-first century, ed. D. Western & M C. Pearl. Oxford University Press, New York & Oxford,

pp. 37—41.

Eiberle, K. (1972). Lebensweise und Bedeutung der Luchses in der Kulturlandschaft. Mammalia Depicta, No. 8. Paul Parey, Hamburg, Berlin.

Fuller, T. K., Berg, W. E. & Kuehn, D. W. (1985). Survival

rates and mortality factors of adult bobcats in north-central

Minnesota. J. Wild!. Manage., 49, 292—6.

Griffith, M. A. & Fendley, T. T. (1986). Pre and post dis­

persal movement behaviour of subadult bobcats of the

Savannah River Plant. In Cats of the World: Biology, Conservation and Management, ed. S. D. Miller & D. D. Everett. National Wildlife Federation, Washington, pp. 277—


Heisey, D. M. & Fuller, T. K. (1985). Evaluation of survival and cause-specific mortality rates using telemetry data. J. Wild!. Manage., 49, 668—74.

IUCN (1988). 1988 IUCN Red List of Threatened Animals.

IUCN, Gland.

Kaczensky, P. (1990) Beaucoup de lynx périssent jeunes. In­

fodienst Wildbiologie, No. 3.

Laurence, W. F. & Yensen, E. (1991). Predicting the impacts

of edge effects in fragmented habitats. Biol. Conserv., 55,


Lovejoy, T. E., Rankin, J. M., Bierregaard, P. 0., Brown, Jr

K. S., Emmons, L. H. & Van der Voort, M. E. (1984).

Ecosystems decay of Amazon forest fragments. In Extinct­

ions, ed. M. H. Nitecki. University of Chicago Press, Chicago, pp. 296—325.

Mallinson, J. (1978). The Shadow of Extinction. Macmillan, London.

Mech, L. D. (1989). Wolf population survival in an area of high road density. Amer. Mid!. Nat., 121, 387—9.

Mech, L. D. & Nelson, E. N. (1990). Evidence of prey-caused mortality in three wolves. Amer. Mid!. Nat., 123, 207—8. Palomares, F., RodrIguez, A., Laffitte, R. & Delibes, M. (1991).

The status and distribution of the Iberian Lynx Felis pardina

in Coto Doñana area, SW Spain. BioL Conserv., 57, 159—69.

Rau, J. R., Beltrán, J. F. & Delibes, M. (1985). Can the in­

crease of fox density explain the decrease in lynx numbers at Doflana? Rev. Ecol. (Terre Vie), 40, 145—50.

Rodriguez, A. & Delibes, M. (1992). Current range and

status of the Iberian lynx Felis pardina, Temminck 1824 in Spain. BioL Conserv., 61, 189—96.

Rogers, P. M. & Myers, K. (1980). Animal distributions,

landscape classification and wildlife management, Coto

Donana, Spain. J. app!. Ecol., 17, 545—65.

Shaffer, M. L. & Samson, F. B. (1985). Population size and

extinction: a note on determining critical population sizes.

Amer. Nat., 125, 144—52.

Soriguer, R. C. (1981). BiologIa y dimámica de una población

de conejos Oryctolagus cuniculus L. en AndalucIa Occidental.

Doñana Acta Vert., 3(3), 1—379.

Trent, T. T. & Rongstad, 0. J. (1974). Home range and sur­ vival of cottontail rabbits in southwestern Wisconsin. J. Wild!. Manage., 38, 459—72.

Vandewalle, Ph. (1989). Le cycle reproducteur de lapin de Garenne Oryctolagus cuniculus en Camargue: influence des factors environnementaux. Gibier Faune Sauvage, 6, 1—25.

База данных защищена авторским правом ©shkola.of.by 2016
звярнуцца да адміністрацыі

    Галоўная старонка