J . Zool., Lond. (1991) 223, 307-321

Phenotypic and genetic characteristics affecting lifetime reproductive success in female Cuvier’s, dama and dorcas gazelles

(Gazella cuvieri, G. dama and G. dorcas)

C . L. ALADOS A N D J. Escos Estacion Experimental de Zonas Aridas (C.S.I.C.),

1, General Segura, 04001, Almeria (Spain)

(Accepted 30 January 1990)

(With 1 figure in the text)

Populations of Cuvier’s, dama and dorcas gazelles were removed from the western Sahara to the Estacion Experimental de Zonas Aridas, Almeria, Spain, between 1970 and 1975. The life-history records of all the females living between 1970 and 1988 were examined.

Results show that fecundity is lower in young females than in middle-aged females, and that juvenile mortality is higher in fawns of young females. The adult body length of Cuvier’s and dorcas gazelles is affected by their early development but this does not apply to the dama gazelle. The adult body length and the birth weight are positively correlated with longevity in female Cuvier’s and dama gazelles, but for the dorcas gazelle only birth weight is related to longevity. Juvenile survival and reproductive performance of the three gazelle species studied are not related to adult body length or birth weight.

A high inbreeding coefficient reduces longevity in Cuvier’s and dama gazelles but not in the dorcas gazelle. The effect of a high inbreeding coefficient is pronounced in the fecundity and juvenile survival of the female dama gazelle, less pronounced but evident in dorcas gazelle. The effects of inbreeding are minimal in Cuvier’s gazelle and are evident only in the twinning rate. This result is surprising in view of the high inbreeding coefficient in Cuvier’s gazelle relative to the other two species.

As in many mammals, birth weights in the three species studied are closely related to the probability of survival during the first days of life. One of the causes of decreasing birth weight in male and female offspring of Cuvier’s gazelles and in female offspring of dorcas gazelles is their inbreeding coefficient. The inbreeding coefficient of the mother also reduces the offspring’s birth weight in both sexes of the dama gazelle, but there is no influence on offspring of Cuvier’s or dorcas gazelles. On the other hand, only the birth weight of the male offspring of the dorcas gazelle is influenced by the birth weight of its mother. Mother’s age and offspring birth weight are related for male offspring of Cuvier’s and dama gazelles.

Contents

Introduction . . . . . . . . . Methods . . . . . . . . . . . Analysis . . . . . . . . . . . Results . . . . . . . . . . .

Longevity . . . . . . . . . Age at first parturition . . . . . Fecundity . . . . . . . . . Juvenile survival . . . . . . .

Conclusions . . . . . . . . . References . . . . . . . . . . .

0952-8369/91/002307+ 15 $03.00

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

Page 308 309 310 31 I 31 I 31 I 312 314 317 319

307 0 1991 The Zoological Society of London

308 C. L. ALADOS AND J . ESCOS

Introduction

Data on fitness are essential for evaluating certain assumptions in modern theories about natural selection. Except for co-operative breeders, the best criterion for fitness is an individual’s lifetime reproductive success, which is a product of variation in lifespan, fecundity and survival of offspring (Fedigan e f al., 1986; Brown, 1988).

In mammals, however, there are few studies measuring lifetime reproductive success and its components, because they take many years (but see Clutton-Brock, Guinness & Albon, 1982; Clutton-Brock et al., 1987; Albon, Clutton-Brock & Guinness, 1987; Clutton-Brock, Albon & Guinness, 1988). Comparable long-term data are available only for some primate species (Altmann et al., 1977; Altmann, 1980; Dunbar, 1980; Sigg et al., 1982; Winkler, Loch & Vogel, 1984; Fedigan e f al., 1986).

A knowledge of life histories is crucial to the study of the population dynamics of a species and to the study of selection. Individual variation is found in parameters directly affecting fitness, e.g. breeding success and survival. Some variations may simply be related to age, for example young and old animals are often less successful than middle-aged animals, while others may be due to the different phenotypical characteristics found in mothers. In captive populations, however, the inbreeding coefficient also affects fitness. Several authors have shown the possible deleterious effects of inbreeding, which generally increases mortality in young animals and reduces fertility in adults (Falconer, 1961; Treus & Lobanov, 1971; Bouman, 1977; Flesness, 1977; Wright, 1977: 44- 96; Lasley, 1978; Seal, 1978; Ralls, Brugger & Ballou, 1979; Senner, 1980; Ralls, Brugger & Glick, 1980).

Close inbreeding is known to have deleterious effects on vigour and fecundity in captive dorcas gazelles (Gazella dorcas) (Ralls, Brugger & Ballou, 1979; Ralls, Brugger & Glick, 1980). Although avoidance of inbreeding appears to be an important factor (Ballou & Ralls, 1982), limitations to outbreeding behaviour are imposed by the cost of migration (Chesser & Ryman, 1986; Waser, Austad & Keane, 1986). Thus if the risk associated with migration is high, it might be more advantageous to remain and breed with relatives. Smith (1979) and Bengtsson (1978) point out that father-daughter mating should occur in polygynous species if the advantages of having a greater proportion of one’s genes represented in one’s offspring are greater than the cost due to inbreeding. The evolution of inbreeding may also be associated with the evolution of altruism (Hamilton, 1975).

Accurate measurements of variables affecting lifetime reproductive success in gazelles have not been made before. The vast majority of the published data concerning reproductive parameters are on Thomson’s gazelle (Gazella thomsoni) (Brooks, 1961; Hvidberg-Hansen, 1970; Robinette & Archer, 1971) and the dorcas gazelle (Jope, 1908; Slaughter, 1971; Ralls, Brugger & Glick, 1980; Bogsch, 1983; Alados, 1984).

Cuvier’s gazelles (Gazella cuvieri), which often bear twins, are noted for their high reproductive rate compared with other gazelles that inhabit savannah areas, such as Gazella subgutturosa from Central Asia. In the mountain areas where Cuvier’s gazelles are found, the vegetation is seasonal and birth numbers show two peaks, one in spring and a lower one in autumn (Olmedo, Escos & Gomendio, 1985), coinciding with the rainy seasons. In some gazelle species, females conceive while lactating and so produce two fawns in one year (Mendelssohn, 1974; Alados, 1984). Baharav (1983) suggests that the timing of births and the ability of suckling females to conceive are not species-specific characteristics but reproductive strategies that depend on daily access to drinking water.

REPRODUCTIVE SUCCESS IN FEMALE GAZELLES 309

Since the gazelles in the present study were in excellent physical condition because of supplementary feeding, the post-partum oestrus, which allows two births a year, was not inhibited in dorcas and Cuvier’s gazelles. The reproductive capability of the dorcas gazelle in which fawns are born singly is not as high as that of Cuvier’s gazelle. Normally, the first fawn is born when the dam is about one and a half years old but a number of females produce their first fawn when one year old (Alados, 1984).

No information exists about the reproductive parameters of the dama gazelle (Guzellu duma), but a report from J. Newby (unpubl.) suggested that they breed in the wild between July and November.

In spite of the fact that the dorcas gazelle is widely distributed, occupying a variety of desert habitats from North Africa to India, several authors have commented on its decline as a result of increased hunting pressure and of agricultural and urban expansion, which appears to be adversely affecting several gazelle habitats (Anderson, 1902; Flower, 1932; Russell, 1949, 195 1; El Negumi, 1952; El Monaiery, 1954; Hoogstraal, 1964; Mendelssohn, 1974; Osborn & Helmy, 1980; Ryder, 1987; Saleh, 1987).

Owing to the recent drastic reduction in numbers of several species of gazelles, there is now an urgent need to protect them and an attempt must be made to understand the intrinsic and extrinsic mechanisms regulating their population density in order to improve their management both in the wild and in captivity. This study is concerned with establishing the phenotypic and genotypic factors affecting components of lifetime reproductive success in female Cuvier’s, dama and dorcas gazelles.

Methods

A founder population of 2 male and 2 female Cuvier’s gazelles, 4 male and 13 female dama gazelles and 36 male and 36 female dorcas gazelles were translocated from the western Sahara to the Estacion Experimental de Zonas Aridas in Almeria (Spain) between 1970 and 1975. The history of the population has been described by Alados, Escos & Vericad (1988) and Escos (In press).

The gazelles are distributed among different herds, each consisting of about 10 individuals, one being an adult male and the remainder adult females with their offspring. The young males are kept with their family groups until they are 4 or 5 months old, when they are placed in bachelor groups.

Each animal is caught the day after birth for marking, weighing and measuring. The identity of its mother is recorded. Body measurements and weight are sometimes recorded when it is adult.

For the present study the life-history records of all the females living in the Estacion Experimental de Zonas Aridas between 1970 and 1988 were examined. All the birth dates of the females and their offspring are known, with the exception of those that arrived as adults.

The following variables have been collected: mother’s age in days on 3 I December 1987 or at the time of death, number of offspring, number of abortions, if the parturition was single or double, interbirth intervals, offspring survival, offspring birth weight, adult body length and the inbreeding coefficients of the mothers and their offspring.

For the purpose of comparing inbred and non-inbred individuals of Cuvier’s gazelles, the animals were divided into 3 categories: non-inbred calves, which consisted of those with an inbreeding coefficient of 0, inbred animals with an inbreeding coefficient greater than 0 but lower than 0.2 and inbred animals with an inbreeding coefficient greater than 0.2. In the case of dama and dorcas gazelles, only 2 categories were considered: non-inbred animals, consisting of those with an inbreeding coefficient of 0, and inbred animals, with an inbreeding coefficient greater than 0. Interbirth intervals were classified into 3 categories: < 200 (2 births a year), 200-300 (1 birth a year) and > 300 days.

For the present study, the following components of lifetime reproductive success were calculated.

310 C. L. ALADOS AND J . ESCOS

Longevity: mother’s age in days at the time of death. Age atfirst birth: mother’s age in days at first live birth. Fecundity: expressed as the number of offspring divided by the reproductive life in years, the reproductive

life being equal to the age at the time of death or on 3 1 December 1987, minus the mean age at first birth. This variable was calculated only for females more than 3 years old, which should have had several breeding attempts.

Abortion rate: the proportion of identified pregnancies which ended in abortions. OfSspring survival to I month: the calves were divided into 2 categories: those that survived 1 month or more

One-month survival rate: the proportion surviving to 1 month of the total offspring born alive to each and those that survived less than 1 month.

mother.

Analysis

Relationships between the mother’s phenotypic characteristics (birth weight and adult body length) and inbreeding coefficient and the components of lifetime reproductive success (longevity, age at first birth, fecundity, abortion rate and one-month survival rate) were investigated by a multiple regression analysis. A linear regression analysis was used to test the associations between birth weight and adult body length. Each component of reproductive success was checked for skew, but only longevity was skewed and required logarithmic transformation.

A xz analysis was used to test for associations between mother’s age and fecundity and offspring survival. A Kolmogorov-Smirnov two-tailed test was used to test the variability of some components of lifetime reproductive success. A Kruskal-Wallis analysis of variance was performed to examine the relationship between inbreeding coefficient, grouped in classes, and the components of lifetime reproductive success. A t-test was used to compare the components of lifetime reproductive success between species.

The hierarchical log-linear model, formulated by Bishop, Fienberg & Holland (1975) and Haberman (1978) was used for the analysis of the interbirth intervals, inbreeding grouped in classes and frequency of twins in Cuvier’s gazelles. These models are useful for uncovering the potentially complex relationships among the variables in a multiway cross-tabulation. In log- linear models all variables that are used for classification are independent variables and the dependent variable is the number of cases in a cell of cross-tabulation.

We examined the relation between birth weight and the inbreeding coefficient of mother and offspring by a one-way ANOVA analysis of variance. In order to remove the effect of repeated sampling of mothers, a factor for mother identity was added in the ANOVA, but no significant effect was observed. Mother’s age was classified into six age classes ranging from class 1 for mothers younger than 2 years old, through 2-year steps to class 6 for mothers older than 10 years. Mother’s age class was added as a factor in the analysis.

To analyse mortality, contingency-table analysis was used followed by logistic regression analysis (Cox, 1970), which relates a binary dependent variable (I-month survival) to any metric independent variable by fitting a logistic curve. In our logistic models, we included the influence of mother and offspring birth weight, sex, mother and offspring inbreeding coefficient, grouped in classes, and age of mother grouped in 3-year age classes. This technique has been used in previous analysis ofjuvenile mortality (Clutton-Brock er al., 1987; Pemberton et al., 1988) and hind fertility (Albon et al., 1986).

Three statistical packages were used in the analysis: SPSS, BMDP and SOLO.

REPRODUCTIVE SUCCESS IN FEMALE GAZELLES 311

Results

Adult body size is often closely related with early development. So there was a significant relationship between the female’s birth weight and her adult body length in Cuvier’s gazelles (r2=0.1 1, F(1,36)=4-35, P<0.05), and in female dorcas gazelles (r2=0.16, F(1,25)=4.86, P= 0.04), but not in female dama gazelles (r2 = 0.001, F( 1,48) = 0.03, N.S.). Since inbreeding coefficient can also affect birth weight, the adult body size, birth weight and inbreeding coefficient will be compared with the components of the lifetime reproductive success by a multiple regression analysis.

Longevi ty

Differences in longevity were pronounced, ranging from 0 to 3834 days in female Cuvier’s gazelles, from 0 to 5027 days in female dama gazelles, and from 0 to 3681 days in female dorcas gazelles. Comparisons of the distribution of longevity between the species showed that dama gazelles live for a significantly longer period (mean = 904 days, S.E. = 175, n = 6 1) than dorcas gazelles (mean=375 days, S.E.=59, n = 133), (t=3.61, d.J= 192, P=0.0004). Cuvier’s gazelles occupy an intermediate position with a mean = 661 days (S.E. = 103, n = 61). The difference is significant between dorcas and Cuvier’s gazelles ( t = 2.56, dLf. = 192, P= 0.01), but not between dama and Cuvier’s gazelles ( t = 1.20, dJ = 120, N.S.). These differences could be associated with a variety of phenotypic or genetic characteristics.

Our first approach to the study of factors associated with longevity was a linear regression analysis of longevity in relation to birth weight. The results for all three species show that gazelles with high birth weights live longer (r2 = 0.09, F ( I ,46) =4.29, P < 0.04 for Cuvier’s gazelle; r2 = 0.45, F(1,68)=56.11, P<O-OOOl for the dama gazelle; r2=0-14, F(1,109)= 17-76, P<O.OOOl for the dorcas gazelle).

Second, a multiple regression analysis was performed with longevity as the dependent variable and birth weight, adult body length and inbreeding coefficient as independent variables, using the smaller sample for which these data were available. For Cuvier’s gazelle adult body length ( t = 12.41, P<O.OOOl) and inbreeding coefficient ( t = -2.10, P=O.O4) are significantly associated with longevity (r2=0-83, F(2,37)=91.17, P<O.OOOl), while birth weight does not improve the model ( t = -0.88, N.S.). Presumably we lost the birth weight effect in the regression analysis because adult body length is explaining the same variation. Similarly, in dama gazelles, the fitted model (r2 = 0.68, F(2,30) = 32-22, P < 0.0001) included adult body length ( t =4.32, P= 0.0002) and inbreeding coefficient ( t = -7.73, P<O*OOO1), while birth weight was excluded ( t = -0.5, N.S.). Finally, there was no relationship between longevity and any of the characteristics studied in dorcas gazelle females (r2=0.05, F(3,23) =0.38, N.S.) (Table I).

Age at first parturition

Between-species comparison of age (in days) at first birth showed that Cuvier’s gazelles bred earliest (mean = 520, S.E. = 26.64, n = 42), followed by dorcas gazelles (mean = 568, S.E. = 15.44, n = 70) and then dama gazelles (mean = 749, S.E. = 49-53, n = 40). The results of t-tests showed a significant difference between Cuvier’s and dama gazelles (1=4.13, d.f. = 80, P=O.OOOI) , and between dorcas and dama gazelles (t=4.25, dJ = 108, P < O.OOOl), but not between Cuvier’s and dorcas gazelles ( I = 1.68, dLf. = 110, N.S.).

312 C. L. ALADOS A N D J . ESCOS

TABLE I F values and signijicance levels fo r multiple linear regression. Differences in the logarithm of the longevity ( y ) were analysed in relation to birth weight. adult body size and inbreeding coeficient

F VI vz r2 1 P Slope

Cuvier’s gazelle Adult body size Inbreeding coefficient Intercept Multiple regression

Dama gazelle Adult body size Inbreeding coefficient Intercept Multiple regression

Dorcas gazelle Birth weight Adult body size Inbreeding coefficient Intercept Multiple regression

12.4 1 -2.10 -2.17

91-17 2 27 0.83

4.32 - 7.73 -0.77

32.22 2 30 0.68

0.50 -0.43 - 1.03

1.97 0.38 3 23 0.05

0~0001 0.05 0.04

0.0002 0.000 1 0.44

0.63 0.67 0.3 1 0.06

0.003 - 1.32 -0.72

0.0002 - 2.90 -0.70

0.17 -0.0009

3.66 - 1.54

The age at which adult female Cuvier’s gazelles first gave birth varied widely between individuals, ranging from 345 days to 1070 days (Kolmogorov-Smirnov two-tailed test, z = 3.28, n = 42, P < 0.001). For female dama gazelles the age at first birth varied from 407 days to 21 36 days (Kolmogorov-Smirnov two-tailed test, z = 3.98, n = 58, P < 0.001). For female dorcas gazelles the age at first birth varied significantly, from 346 to 929 days (Kolmogorov-Smirnov two-tailed test,

Multiple regression analysis of birth weight, adult body length and inbreeding coefficient in relation to age at first birth showed no significant associations in Cuvier’s gazelles (multiple regression r2=0*09, F(3,31)= 1.09, N.S.) or in dorcas gazelles (r2=0.21, F(3,16)= 1.44, N.S.). However, in dama gazelles age at first birth was significantly associated with inbreeding coefficient (u2=0.19, F(1,35)=8.48, P=0.006) but not birth weight ( t = 1.76, N.S.) or adult body length ( I = 1.23, N.S.). Dama gazelle females with high inbreeding coefficients breed later.

Z= 1.96, n=70, P=O.OOl).

Fecundity

Before an analysis of fecundity could be performed, it was necessary to allow for the possible effects of a female’s age on fecundity. Yearling Cuvier’s gazelles have a lower fecundity than animals of two years and older. However, if only females aged between two and 10 years are considered, no age-related differences in fecundity are apparent ( x 2 = 13.3, d$ = 8, N.S.). For similar reasons, only female dama gazelles between two and 1 1 years old were considered (x2 = 4.1, d,$ = 8, N.S.), and for female dorcas gazelles, only individuals between one and eight years old were considered because yearlings are just as fecund as older females (x2=0.43, d$ =6, N.S.).

The number of offspring per year ranges from 0 to 3.24 in female Cuvier’s gazelles. Differences in fecundity were not significantly correlated with adult body length, birth weight or inbreeding coefficient (multiple regression r2=0.14, F(3,17)=0.96, N.S.).

REPRODUCTIVE SUCCESS IN FEMALE GAZELLES 313

In the 48 female dama gazelles studied, fecundity ranged from 0 to 1.86 offspring annually, but again neither birth weight, adult body length nor inbreeding coefficient was associated with this variation (multiple regression r2=0.09, F(3,29) =0.93, N.S.). Sample size for the multiple regression analysis was limited, because although the inbreeding coefficient is known for all the individuals, only a few of them were measured as adults. Thus we conducted a separate analysis using inbreeding coefficient grouped into classes. We found inbreeding coefficient was significantly associated with fecundity (Kruskal-Wallis analysis of variance xz = 7.59, d.J = 1, n = 48, P = 0.006).

Female dorcas gazelle fecundity varied from 0 to 2.18 calves per year. No associations were observed with birth weight, adult body length or inbreeding coefficient (multiple regression rZ=0.03, F(3,10)=0.1, N.S.).

Comparisons between the three species revealed that the annual number of offspring for each female Cuvier’s gazelle was significantly higher (mean = 1.56, S.E. = 0.17, n = 27) than for the dama gazelle (mean = 0.91, S.E. = 0.07, n = 48) ( t = 3.92, d.J = 73, P= 0.0002). Differences between dama and dorcas (mean = 1.3 1, S.E. = 0.08, n = 50) gazelles were also significant ( t = 3.63, d.J = 96, P=0.0005), but not between Cuvier’s and dorcas gazelles ( t = 1.48, d.J = 75, N.S.).

In spite of the fact that no relationship exists between fecundity and inbreeding in Cuvier’s gazelle, a negative relationship was found between inbreeding coefficient and the frequency of twins. In order to test that interbirth intervals were not influenced by the twinning rate, a hierarchical log-linear/backward analysis was performed for the three variables (interbirth intervals and inbreeding grouped in classes, and frequency of twins), with all the variables used for classification as independent variables. The results showed that the best model generating classes (i.e. the model providing the best fit) was inbreeding vs. frequency of twins, with a likelihood ratio x2 = 8.76, d.J = 2, P = 0.01. There were no significant associations between inbreeding and interbirth intervals (likelihood ratio x2 = 7.93, d.J =4, P= 0.09) or between twinning rate and

TABLE I1 Logistic regression of juvenile survival and offspring characteristics (birth weight, sex and inbreeding coefi-

cient)

Cuvier’s gazelle Birth weight Sex Inbreeding Model

Dama gazelle Birth weight Sex Inbreeding Model

Dorcas gazelle Birth weight Sex Inbreeding Model

23.39 1 2.98 1 0.03 1

35.59 3

56.84 I 20.3 1 0.23 1

108.54 3

53.9 1 0.02 1 1.33 1

75.02 3

0.000 1 N.S. N.S. 0.000 1

0.000 1 0.001

N.S. 0.0001

0.001 N.S. N.S.

0~0001

0.10 0.01 0.00 0.16

0.18 0.07 0.00 0.30

0.12 0.00 0.00 0.16

3 14 C. L. ALADOS A N D J . ESCOS

interbirth intervals (likelihood ratio x2 = 5.69, d.f. = 2, P= 0.058). In summary, the higher the inbreeding coefficient, the fewer pairs of twins are born, but interbirth intervals are unaffected.

Juvenile survival

Ofspring characteristics afecting juvenile survival

On average, 16.1% of the Cuvier's fawns born in the Estacion Experimental de Zonas Aridas died during the first month of life. Seventy-two percent of the Cuvier's fawns dying in the first three months of life died within a week of birth. Offspring survival was analysed as a binary variable. This dependent variable was related to metric or categorical variables by a logistic regression. In the Cuvier's gazelle population, offspring survival was the dependent variable with offspring birth weight, sex and inbreeding coefficient (grouped in classes) as independent variables. Birth weight was the only significant variable in predicting juvenile survival (Table 11).

As predicted by the logistic regression, in Cuvier's gazelles, mean birth weights were higher for individuals surviving beyond one month than for those dying within a month of birth (2-832 kg and 1.583 kg, respectively, for females; 2.844 kg and 2.220 kg, respectively, for males).

Juvenile mortality of dama gazelles was higher than that of Cuvier's gazelles. Of 189 males and 183 females born, the proportion dying in the first month of life was 32.8% and 27.3%, respectively. Logistic regression (Table 11) shows that birth weight is the most important factor in predicting juvenile survival, with sex of offspring the second. Offspring inbreeding coefficient was not significantly associated with survival to one month. Thus, mean birth weights were higher for

i 1

&YiU G. cuvieri b i OXUl G. darna

/ G. dorcas /

I

, --- - - I :/

0.00 --=------ 3 n n n u u u U I I I I I 0 1 2 3 4 5 6 7 8

Birth weight (kg)

FIG. 1. Logistic curve fitted to the distribution of probabilities ofjuvenile survival on birth weight for males and females of Cuvier's y = l / ( l +exp(5.27-2.73~)), dama y= l / ( l +exp(8.71-2.08~)) and dorcas y = l/(l fexp(5.12-4.09~)) gazelles. The proportion of offspring surviving represented by each symbol was calculated as a percentile of the weight.

REPRODUCTIVE SUCCESS IN FEMALE GAZELLES 315

individuals surviving beyond one month than for those dying within a month of birth (4.62 kg and 3.75 kg, respectively, for females; 5.25 kg and 4.25 kg, respectively, for males).

In dorcas gazelles, 3 1 of 182 males (17%) and 41 of 182 females (28%) died within the first month of life. Logistic regression (Table 11) shows that of the variables studied, birth weight is the only variable affecting juvenile survival. Offspring inbreeding coefficient and sex does not improve the model. The mean birth weight was higher for individuals surviving beyond one month than for those dying within a month of birth (1.63 1 kg and 1.426 kg, respectively, for females; 1.762 kg and 1.436 kg, respectively, for males).

Figure 1 shows the probability ofjuvenile survival plotted against birth weight for both sexes of Cuvier’s, dama and dorcas gazelles. It can be seen that, for the three species studied, birth weight is related to juvenile survival.

Low birth weights could be due to high inbreeding coefficients. To test this, birth weight was compared with the inbreeding coefficient grouped into classes. A mother’s age factor was added in the ANOVA to remove the effect of this intervening variable. A significant relationship was observed for both male (ANOVA analysis of variance F= 11.23, d-f. = 1, P=O.OOl) and female Cuvier’s gazelles (ANOVA analysis of variance F= 13.05, d$ = 1, P = 0.001). In both cases the highest mean birth weight fell within the inbreeding coefficient range >0-0.2> (2.765 for females, 2.812 for males). The lowest mean birth weight was found in animals with the highest inbreeding coefficient (> 0.2) (2.418 for females, 2.541 for males). In dama gazelles the inbreeding coefficient did not affect birth weight of either males or females (ANOVA analysis of variance F=0-06, d.f. = 1, N.S., for males; F=0.14, d$= 1, N.S., for females). In dorcas gazelles male birth weight was not affected by the inbreeding class (ANOVA analysis of variance F=0.54, d-f. = 1, N.S.), but female birth weight was (ANOVA analysis of variance F=4.94, d-f. = 1, P=O.O3).

Maternal characteristics afecting juvenile survival

Cross-species comparison of offspring survival to one month for each female gazelle showed that the one-month survival rate of offspring is higher for Cuvier’s gazelle (mean=0.69, S.E.=0.05, n=41) than for thedamagazelle (mean=0-56, S.E.=O-04, n=73)(t = 1.82, d.J= 112, P=O.O7). The dorcas gazelle also has a higher offspring survival to one month (mean=0.65, S.E. = 0.04, n = 96) than the dama gazelle, but the difference is not significant ( t = 1.60, d$ = 167, N.S.). Similarly, no difference was observed between dorcas and Cuvier’s gazelles (t =0.53, d.J = 135, N.S.).

The rate of abortions has been analysed in relation to maternal characteristics (birth weight, adult body length and inbreeding coefficient). The results of the multiple regression analysis for females over three years old are not significant for either Cuvier’s (r2 =0.08, F(3,32) =0.89, N.S.), dama (r2=0.03, F(3,36)=0.42, N.S.) or dorcas gazelles (r2=0.01, F(3,16)=0.07, N.S.).

An analysis was made by logistic regression of the relationship between one-month juvenile survival and maternal characteristics. One-month offspring survival was related as a dependent binary variable with the mother’s age class, mother’s inbreeding class, mother’s birth weight and offspring sex as independent variables. The results for Cuvier’s gazelle show that the model providing the best fit includes mother’s age and offspring sex, while mother’s inbreeding coefficient and mother’s birth weight do not improve the fit of the model to the data. In dama and dorcas gazelles, logistic analysis suggests that mother’s age class and mother’s inbreeding coefficient are significantly associated with juvenile survival. The other two variables, mother’s birth weight and offspring sex, do not improve the model (Table 111).

316 C. L. ALADOS A N D J . ESCOS

TABLE 111 Logistic regression of juvenile survival and mother characteristics (age, inbreeding, birth weigh1 and off-

spring sex)

x2 drf. P 9

Cuvier’s gazelle Age 7.93 I 0.005 0.04 Inbreeding 0.59 I N.S. 0.00 Birth weight 0.22 I N.S. 0.00 Offspring sex 6.02 1 0.01 0.03 Model 13.56 4 0.009 0.07

Dama gazelle Age 10.08 I 0.001 0.04 Inbreeding 8.62 1 0.003 0.04 Birth weight 1.45 1 N.S. 0.00 Offspring sex 1.74 1 N.S. 0.01 Model 23.03 4 09001 0.08

Dorcas gazelle Age 10-51 I 0.001 0.03

Birth weight 0.78 1 N.S. 0.00 Offspring sex 0.02 1 N.S. 0.00 Model 18.21 4 0.001 0.06

Inbreeding 4.30 1 0.04 0.0 1

TABLE IV Comparisons ofjuvenile survival amongst mother ’.s age classes

Mother’s age classes (years)

0-3 3-6 6-9 < 9

Cuvier’s gazelle Survival longer than 1 month (n) % Total (n)

x2 = 6.47, d$ = 2, P= 0.04

Dama gazelle Survival longer than 1 month ( n ) %I

Total (n ) x2= 18.30, d J = 3 , P=0,0004

Dorcas gazelle Survival longer than 1 month (n ) %I

Total ( n ) x2=8.88, drf.=3, P=0.03

76 68.5

I l l

48 52.2 92

151

214 70.6

54 78.3 69

89 72.9

122

122

157 77.7

27 90.0 30

46 25 78.3 83.3 60 30

59 1 1 86.8 64.7 68 17

REPRODUCTIVE SUCCESS IN FEMALE GAZELLES 317

Since mother’s age affects juvenile survival in the three species, a more detailed study was carried out (Table IV). For female Cuvier’s gazelles less than three years old, 68.5% of their offspring survive more than a month, while for females more than three years old 84.1% survive. No significant difference exists between age and juvenile survival in mothers more than three years old (x2 = 3.2, d$ = 1, N.S.). For female dama gazelles less than three years old, juvenile survival is also low (52.2%) and increases to 83.3% for females over nine years old. However, after three years, no association was found between the mother’s age and offspring survival to one month (x2 = 1.68, d$ = 2, N.S.). There is a similar association between mother’s age and juvenile survival in the dorcas gazelle.

In order to see if a mother’s inbreeding coefficient reduces her offspring birth weight, the effects of mother’s inbreeding coefficient on birth weights of male and female offspring were analysed separately, and the gazelles were classified into six age classes in order to remove the effect of a intervening variable, adding a factor for mother’s age in the ANOVA. In Cuvier’s gazelle mother’s inbreeding coefficient does not affect offspring birth weight in either males (ANOVA analysis of variance F= 2.3 I , d$ = 2, N.S.), or females (ANOVA analysis of variance F= 1.64, d$ = 2, N.S.). Similarly, in dorcas gazelles mother’s inbreeding coefficient does not influence the birth weight of either males (ANOVA analysis of variance F=0.05, d$ = 1, N.S.) or females (ANOVA analysis of variance F=0.79, d$ = 1, N.S.). However, in the dama gazelle, mother’s inbreeding coefficient reduces birth weight in both male (ANOVA analysis of variance F= 8.68, d$ = 1 , P < 0.004) and female offspring (ANOVA analysis of variance F= 10.03, d$ = 1 , P = 0.002). Mother’s birth weight does not affect offspring birth weight of either males (linear regression r2 = 0.01, F( l,79) = 0.9, N.S.) or females (r2=0.04, F(1,68) = 3.2, P= 0.08) in either Cuvier’s gazelles or dama gazelles (linear regression 9 = 0.01, F( 1,69) = 0.5, N.S., for males; r2 = 0.01 F( 1,64) = 0.6, N.S. for females). Female dorcas gazelle birth weights are not associated with their mother’s birth weight (multiple regression r2=0.001, F(1,109)=0.11, N.S.), but a positive relationship was found between the birth weight of male offspring and the birth weight of their mothers (3 = 0.04, F( 1,118) = 4.63, P = 0.03).

The association between the most important characteristics affecting juvenile survival (mother’s age and offspring birth weight) are only significant for male offspring of Cuvier’s (ANOVA analysis of variance, F(2,82) = 4.34, P= 0.02) and dama (F(3,95) = 3.26, P= 0.02) gazelles after controlling for the effect of inbreeding.

Conclusions

Female longevity is one component of female lifetime reproductive success and, for both Cuvier’s and dama gazelles, is positively related to birth weight and adult body length, and negatively influenced by increasing inbreeding coefficient. In dorcas gazelles, only birth weight is correlated with longevity.

In mammals, mother’s age can also affect other components of lifetime reproductive success such as fecundity and juvenile survival (Turner & Dolling, 1965, in domestic sheep, Ovis aries; Nievergelt, 1966, in C a p o ibex; Drickamer, 1974, in Macaca mulutta; Clutton-Brock et al., 1982, 1987, in Cervus eluphus). The three gazelle species studied have a lower fertility rate at high and low ages, and the mother’s age has an effect on juvenile survival.

Variation in birth weight is closely related to juvenile survival in red deer (Guinness, Clutton- Brock & Albon, 1978; Albon et al., 1987; Clutton-Brock et al., 1987) and may also influence adult body size and reproductive performance (Doney & Gunn, 1981; Bongaarts & Potter, 1983;

318 C . L. ALADOS AND J. ESCOS

Clutton-Brock et al., 1987). In female Cuvier’s and dorcas gazelles the birth weight affects adult body length but this is not the case with the dama gazelle.

Although other authors, working with red deer, have observed a positive relationship between a female’s weight and her fertility (Mitchell & Brown, 1974), this could be due to the effect on reproduction of body condition rather than body size, as Clutton-Brock et al. (1982) have suggested. Albon el al. (1986) demonstrated that body condition, independent of body weight, is an important factor affecting fertility in red deer. In that study, fertility was inversely related to skeletal size, since at the same body weight a skeletally small hind has a greater muscle mass than a skeletally large hind.

Since gazelles at the Estacion Experimental de Zonas Aridas are in excellent physical condition because of supplementary feeding, body size is not a reflection of their body condition but probably of inheritance, which may explain why, in none of the three gazelle species studied, does adult body length or birth weight influence the age at first birth, fecundity, offspring survival or abortion rate.

Many authors have reported that a high inbreeding coefficient reduces fertility and juvenile survival (Falconer, 1961; Treus & Lobanov, 1971; Bouman, 1977; Flesness, 1977; Wright, 1977: 44-96; Lasley, 1978; Seal, 1978; Ralls, Brugger & Ballou, 1979; Senner, 1980; Ralls, Brugger & Glick, 1980), and the characters most severely affected by inbreeding are those expressed early in life, such as perinatal mortality and birth weight (Falconer, 1961; Lasley, 1978).

Theoretically, species that naturally inbreed in the wild should show less of a deleterious effect when subjected to inbreeding in captivity, but little is known about the extent to which ungulates normally inbreed in the wild, although several models suggest that inbreeding could be favoured in polygynous systems (Waser et al., 1986). Our results show that inbred female dama gazelle produce fewer offspring than non-inbred females, as Bouman (1977) showed in Przewalski horses (Equusprzewalskii). The high inbreeding coefficient also leads to a reduction in juvenile survival in dama gazelle, as in many other captive ungulates (Falconer, 1961; Wright, 1977: 44-96; Lasley, 1978; Ralls, et al., 1979,1980; Ryder, 1987). The influence of inbreeding on the age at first breeding is not clear-see Ralls et al. (1980)-but in the dama gazelle the high inbreeding coefficient also decreases fertility by raising the age at first conception. The rate of abortions is not related to the mother’s inbreeding coefficient in dama gazelles. The fecundity of dorcas gazelles is also affected by a high inbreeding coefficient, but not as much as in dama gazelles, probably because of the higher number of founders of our population. As in the dama gazelle, offspring survival is affected by the high inbreeding coefficient. However, the Cuvier’s gazelle, with a higher level of inbreeding than the other two species, is affected only in the frequency of twinning.

We suggest that the influence of the inbreeding coefficient on lifetime reproductive success is less in Cuvier’s gazelle than in the other two species because of ecological factors. Cuvier’s gazelle lives in mountain habitats, in smaller social groups than the other two species (Sclater & Thomas, 1898: 109-1 14), which are desert gazelles and have to migrate to find free water and food (Newby, unpubl. and 1984). Migrations involve large numbers, and although dama gazelles normally live in groups of up to 10 individuals, the migrating herd may number over 100 (Newby, 1984). According to theory, species that are naturally inbred in the wild should show less of a deleterious effect when subjected to inbreeding in captivity (Ballou & Ralls, 1982).

Birth weight, which is considered to be one of the variables most closely related to the chances of juvenile survival in ungulates (Guinness et al., 1978; Doney & Gunn, 1981; Albon et al., 1987; Clutton-Brock el al., 1987; this study), is controlled partly by genetic and partly by environmental factors. An important genetic factor that influences the offspring birth weight is the inbreeding

REPRODUCTIVE SUCCESS IN FEMALE GAZELLES 319

coefficient. Similar results were observed on the birth weight of Speke’s gazelle (Gazella spekei) (Ryder, 1987).

We thank the persons who have collected gazelle data since 1971, and the Estacion Experimental de Zonas Aridas for providing free access to the gazelle registry. Sebastian Vidal helped us with the computer, Mari Carmen Cazorla typed the draft and Rosalind Corrigan and Susan Eltringham corrected the English. We also thank Dr K. Eltringham and three anonymous referees for comments on earlier drafts of this paper.

REFERENCES

Alados, C. L. (1984). La reproduccion en Gazella dorcas. Doiiana Acta uert. ll(2): 243-261. Alados, C. L., Escos, J. & Vericad, J. R. (1988). Captive populations of northwest African Antilopinae and Caprinae at the

Estaci6n Experimental de Zonas Aridas. In Conservation and biology of desert antelopes: 199-21 1. Dixon, A. & Jones, D. (Eds). Bromley: Christopher Helm Ltd.

Albon, S. D., Clutton-Brock, T. H. & Guinness, F. E. (1987). Early development and population dynamics in red deer. 11. Density-independent effects and cohort variation. J . Anim. Ecol. 5 6 69-81.

Albon, S. D., Mitchell, B., Huby, 8. J . & Brown, D. (1986). Fertility in female Red deer (Ceruus elaphus): the effects of body composition, age and reproductive status. J . Zool., h n d . ( A ) 209 447-460.

Altmann, J . (1980). Baboon mothers and infants. Cambridge, Mass. & London: Harvard University Press. Altmann, J. , Altmann, S. A,, Hausfater, G. & McCuskey, S. A. (1977). Life history of yellow baboons: physical

Anderson, J. (1902). Zoology ofEgypt: Mammalia. (Revised and completed by W. E. de Winton). London: Hugh Rees. Baharav, D. (1983). Reproductive strategies in female Mountain and Dorcas gazelles (Gazella gazella gazella and Gazella

Ballou, J. & Ralls, K. (1982). Inbreeding and juvenile mortality in small populations of ungulates: a detailed analysis. Biol.

Bengtsson, B. 0. (1978). Avoiding inbreeding: at what cost? J . rheor. Biol. 73: 439-444. Bishop, Y. M. M., Fienberg, S. E. & Holland, P. W. (1975). Disc~ete~ul t ivar jafe analysis. Cambridge, Mass.: MIT Press. Bogsch, 1. (1983). Zuchtergebnisse bei Dorkasgazellen, Gazella dorcas (L.), im Zoo Budapest. Zool. Gart., (N.F.) Jena

Bongaarts, J. & Potter, R. G. (1983). Fertility, biology and behauiour. New York: Academic Press. Bouman, J. (1977). The future of Przewalski horses Equus przewalskii in captivity. Int. Zoo Yb. 17: 62-68. Brooks, A. C. (1961). A study of the Thomson’s gazelle (Gazella thomsonii Giinther) in Tanganyika. Colon. Res. Publs

No. 25: 1-147. H.M.S.O., London. Brown, D. (1988). Components of lifetime reproductive success. In Reproductive success. Studies of individual variation in

contrasting breeding systems: 439453. Clutton-Brock, T. H. (Ed.). Chicago: University of Chicago Press. Chesser, R. K. & Ryman, N. (1986). Inbreeding as a strategy in subdivided populations. Evolution. Lawrence, Kans. 40(3):

6 16-624. Clutton-Brock, T. H., Albon, S. D. & Guinness, F. E. (1988). Reproductive success in male and female red deer. In

Reproductive success. Studies of indicidual uariation in contrasting breeding systems: 325-343. Clutton-Brock, T. H. (Ed.). Chicago: University of Chicago Press.

Clutton-Brock, T. H., Guinness, F. E. & Albon, S. D. (1982). Red deer. Behauiour and ecology of two sexes. Edinburgh: Edinburgh University Press.

Clutton-Brock, T. H., Major, M., Albon, S. D. & Guinness, F. E. (1987). Early development and population dynamics in red deer. 1. Density-dependent effects on juvenile survival. J . Anim. Ecol. 56: 53-67.

Cox, D. R. (1970). The analysis of binary data. London: Methuen. Doney, J. M. & Gunn, R. G. (1981). Nutritional and other factors in breeding performance of ewes. In Environmental

Drickamer, L. C. (1974). A ten-year summary of reproductive data for free-ranging Macaca mularra. Folia primatol. 21:

Dunbar, R. I. M. (1 980). Demographic and life history variables of a population of gelada baboons (Theropithecusgelada).

El Monaiery, A. M. (1954). Situation of wild mammals in Egypt. Bull. 2001. Soc. Egypt No. 12: 1. El Negumi, A. (1952). (Egyptian mammals.] Alexandria: Khalil Ibrahim. [In Arabic.]

development, reproductive parameters, and infant mortality. Primates 1 8 315-330.

dorcas). J . Zool., Lond. 200: 445453.

Conserv. 2 4 239-272.

53: 309-312.

factors in mammal reproduction: 167-177. Gilmore, D. & Cook, 9. (Eds). London: Macmillan.

61-80.

J . Anim. Ecol. 4 9 485-506.

320

Escos, J. (In press). Gazella cuvieri studbook Edmi gazelle 1989. Almeria: Instituto de Estudios Almerienses Falconer, D. S. (1961). Quantiratire genetics. New York: Ronald Press. Fedigan, L. M., Fedigan, L., Gouzoules, S., Gouzoules. H. & Koyama, N. (1986). Lifetime reproductive success in female

Flesness, N. R. (1977). Gene pool conservation and computer analysis. fnt. Zoo Yb. 17: 77 81. Flower, S. S. (1932). Notes on the Recent mammals of Egypt, with a list of the species recorded from that kingdom. Proc.

Guinness, F. E., Clutton-Brock, T. H . & Albon, S. D. (1978). Factors affecting calf mortality in red deer (Crrzws rhphus).

Haberman, S. J . (1978). Analysis of qualitatii~e data 1 . New York: Academic Press. Hamilton, W. D. (1975). Innate social aptitudes of man: an approach from evolutionary genetics. In Biosocial

Hoogstraal, H. (1964). A brief review of the contemporary land mammals of Egypt (including Sinai). 3. Carnivora,

Hvidberg-Hansen, H. ( I 970). Contribution to the knowledge of the reproductive physiology of the Thomson's gazelle

Jope, M. (1908). Meine Erfahrungen in der Aufzucht von Dorcas-gazellen und Axis- und Schweinshirschen. Zool. Beob.

Lasley, J . F. (1978). Genetics of lirestock improvement. Englewood Cliffs, N J : Prentice-Hall. Mendelssohn, H . (1974). The development of the populations of gazelles in Israel and their behavioural adaptations. IUCN

Mitchell, B. & Brown, D. (1974). The effects of age and body size on fertility in female red deer (Cercus elaphus L.). Proc.

Newby, J . (Unpublished). The ecological resources qf'the Ouadi Rime-Ouadi Achim Reserue Faunal Reseri,e. Chad. Rome:

Newby. J. E. (1984). Large mammals. In Saharu desert: 177-190. Cloudsley-Thompson. J . L. (Ed.). Oxford: Pergamon

Nievergelt, B. (1966). Unterschiede in der Setzzeit beim Alpensteinbock (Capra ibex L.). Rev. Suisw Zool. 73: 446-454. Olmedo, G., Escos, J. & Gomendio, M. (1985). Reproduction de Gazella cucieri en captivite. Mammalia 4 9 501-507.

C . L . A L A D O S A N D J . ESCOS

Japanese macaques. Folia primatol. 47: 143- 157.

zool. Soc. Land. 1932: 369-450.

J . Anim. Ecol. 47: 817-832.

anthropology: 133-155. Fox, R. (Ed.). London: Malaby.

Hyracoidea, Perissodactyla, and Artiodactyla. J . Egypt Puhl. Hlrh Assoc. 39: 205-239.

(Gazella thomsoni Gunther). Mammalia 3 4 55 1-563.

4 9 374-376.

Publs (N.S.) 2 4 722-743.

Int. Congr. Game Biol. 11: 89-98.

FAO. (Unpubl. report 1978).

Press.

Osborn, D. J . & Helmy, I. (1980). The contemporary land mammals of Egypt (including Sinai). Fieldiana (Zool.) N.S. No. 5: 1 579.

Pemberton, J . M., Albon, S. D., Guinness, F. E., Clutton-Brock, T. H. & Berry, R. J . (1988). Genetic variation and juvenile

Ralls. K.. Brugger, K. & Ballou, J. (1979). Inbreeding and juvenile mortality in small populations of ungulates. Science,

Ralls, K., Brugger, K. & Glick, A. (1980). Deleterious effects of inbreeding in a herd of captive dorcas gazelle Ga:ella

Robinette, W. L. & Archer, A. L. (1971). Notes on ageing criteria and reproduction ofThomson's gazelle. E. A,fr. Wildl. J .

Russell, T. (1949). Desert fauna. Bull. zoo/. Soc. Egypt No. 8: 5-8. Russell, T. (1951). The fauna of the Egyptian deserts. Bull. zool. Soc. Ezypt No. 9: 19 21. Ryder, 0. A. (1987). Conservation action for gazelles: an urgent need. Trends E c d . Evol. 2: 143 144. Saleh, M. A. (1987). The decline of gazelles in Egypt. B i d . Conserc. 3 9 83-95. Sclater, P. L. & Thomas, D. ( I 898). The hook ofantelopes 3. London: Porter. Seal, U . S. (1978). The Noah's Ark problem: multigeneration management ofwild species in captivity. In Endangeredbirds:

management techniques , f . r preserring threutmed species: 303 3 19. Temple, S. A. (Ed.). Madison: University of Wisconsin Press.

Senner, J . (1980). Inbreeding depression and the survival of zoo populations. In Conservation biology: an evolutionary- ecological perspective: 209-224. Soule, M. E. & Wilcox, B. A. (Eds). Sunderland, Mass.: Sinauer Associates.

Sigg, H., Stolba, A,, Abegglen, J.-J. & Dasser, V. (1982). Life history of hamadryas baboons: physical development. infant mortality, reproductive parameters and family relationships. Primates 2 3 473- 487.

Slaughter, L. (1971). Gestation period of the dorcas gazelle. J . Mammal. 52: 480-481. Smith, R. H. (1979). On selection for inbreeding in polygynous animals. Heredity, Lond. 4 3 205-21 1. Treus, V. D. & Lobanov, N. V. (197 I). Acclimatisation and domestication of the eland Taurotragus or?.\- at Askanya-Nova

survival in red deer. Ecolution. Lawrence, Kans. 42(5): 921-934.

Wash. 206 1101-1 103.

dorcas. Inr. Zoo Yb. 2 0 137-146.

9 83-98.

Zoo. In t . Zoo Yb. 11: 147-156.

R E P R O D U C T I V E SUCCESS I N F E M A L E G A Z E L L E S 32 1

Turner, H. N. &Dolling, C. H. S. (1965). Vital statistics for an experimental flock of merino sheep. 11. The influence of age

Waser, P. M., Austad, S. N. & Keane, B. (1986). When should animals tolerate inbreeding? Am. Nut. 128(4): 529-537. Winkler, P., Loch, H. & Vogel, C. (1984). Life history of Hanuman langurs (Presbytis entellus). Reproductive parameters,

Wright, S. (1977). Evolution and the genetics of populations 3 . Experimental results and erolutionary deductions. Chicago:

on reproductive performance. Aust. J . ugnc. Res. 1 6 699-7 12.

infant mortality, and troop development. Folia primatol. 43 1-23.

University of Chicago Press.