Relationship between genetic diversity in the nematode Trichostrongylus colubriformis and breeding management in ten dairy-goat farms

Relationship between genetic diversity in the nematode Trichostrongylus colubriformis and breeding management in ten dairy-goat farms

veterinary parasitology ELSEVIER Veterinary Parasitology 66 (1996) 213-223 Relationship between genetic diversity in the nematode Trichostrongylus c...

586KB Sizes 0 Downloads 38 Views

veterinary parasitology ELSEVIER

Veterinary Parasitology 66 (1996) 213-223

Relationship between genetic diversity in the nematode Trichostrongylus colubriformis and breeding management in ten dairy-goat farms G. N'Zobadila a, N. Gasnier b, j. Cabaret b,. a MNHN, Laboratoire de Biologie parasitaire, Protozoologie, Helminthologie, 61 rue de Buffon, 75231 Paris Cedex 05, France b INRA, Station de Pathologie aviaire et de Parasitologie, Ecologie des Parasites, 37380 Nouzilly, France

Received 13 November 1995; accepted 26 February 1996

Abstract Ten dairy-goat farms were investigated in center-west of France for genetic variability of Trichostrongylus colubriformis in relation to breeding management. Farm management data were obtained from a questionnaire. Genetic variability was based on two polymorphic enzymes, malate dehydrogenase (MDH) and glucose-phosphate isomerase (GPI). After their establishment, the farms were subsequently isolated from introduction of strongyle worms as shown in the questionnaire; this was also suggested by the absence of a relationship between genetic variability and distance between farm locations. The genetic variability which was recorded could then be ascribed in part to the influence of management. The breeding management estimates combined the fact that animal breeding was the main economic resource; that goats were or were not the only animal bred; and that there was or was not free access to exercise yards in winter. The farms that were similar on the basis of breeding management, were also similar in the frequency of allozymes, indicating that the chosen allozymes were not neutral in respect to environment. Genetic variability was not related to the frequency of T. colubriformis in the strongyle community, this being possibly due to the fact that our farm samples predominantly harboured T. colubriformis. Between-farm genetic variability was positively correlated to the size of herd

* Corresponding author. 0304-4017/96/$15.00 Copyright © 1996 Elsevier Science B.V. All fights reserved. PII S 0 3 0 4 - 4 0 1 7 ( 9 6 ) 0 1 0 0 9 - 6

214

G. N'Zobadila et al. / Veterinary Parasitology 66 (1996) 213-223

(P < 0.01), probably due to the fact that larger herds were originally constituted from several different herds. Keywords: Nematode; Trichostrongylus colubrifi~rmis; lsoenzyme;Dairy-goat;Breedingmanagement

1. Introduction

Biochemical and molecular genetic studies have revealed that most natural populations are genetically heterogeneous (Powell, 1975; Nadler, 1990). Much debate, as reviewed in Cain and Provine (1991), has focused on how this heterogeneity is maintained. Selection (see Bryant, 1974 on arthropods and rodents) was opposed to neutral mutation (mutation not associated with fitness in a specified environment) and genetic drift (fluctuation by chance in small populations). The use of enzyme separative techniques in digestive-tract nematodes (Andrews and Beveridge, 1990) has proved useful in the characterization of experimentally reared strains and natural populations of Teladorsagia circumcincta (Gasnier et al., 1992) and Trichostrongylus colubriformis (N'Zobadila et al., 1993). The observed differences between experimentally reared strains and natural populations are an example of genetic variability related to environment, indicating that isoenzymes were not selectively neutral. Environment includes climate, host species and their management. T. colubriformis is found in very different climatic areas and in various hosts (sheep, goats and cattle). It is also found, within a region, in all dairy-goat farms, indicating a wide range in adaptation to breeding management. Although present in all farms, T. colubriformis frequency in the worm communities varies greatly, indicating that farms are more or less favourable for this particular strongyle. Dairy-goat farm management is very diversified in the west central area of France (Cabaret et al., 1989) and could thus result in T. colubriformis in higher or lower frequencies in the community of worms harboured by goats. Diverse breeding-management could also result in heterogeneity between populations of T. colubriformis, that could be shown by means of isoenzymatic studies. The main characteristics of goat breeding management are the variable intensity of use of pastures and isolation of helminthic populations on given farms. The intensity of use of pastures is logically correlated to intensity of infection (Cabaret et al., 1989), as the free-living stages of parasites are located on pastures. Helminthic isolation in dairy-goat farms is a clear fact: herds are constituted at the origin of farms and further animal introductions consists of (i) young uninfected goats bred in zero-grazing conditions (ii) or bucks that originated from zero-grazing farms or pasture grazing farms and remained indoors after introduction, thus eliminating chances of helminth introduction (Cabaret and Gasnier, 1994). The description of management cannot be restricted to pasture utilization or animal introductions, and should include a large range of variables (size of herd, duration of kidding period among others), to be really effective in discriminating between the farms (Cabaret and Gasnier, 1994). We may then ask if the T. colubriformis found in the different farms are genetically similar, and if they are not, which breeding management characteristics may influence genetic variability. The objectives of the present work were to: (i) estimate the range of

G. N'Zobadila et al. / Veterinary Parasitology 66 (1996) 213-223

215

allozyme variation in T. colubriformis from different farms located in the same area, and (ii) relate genetic diversity as recorded from allozyme frequencies to the frequency of T. colubriformis in each farm worm community and to characteristics of breeding management.

2. Materials and methods

Ten farms were selected from previous unpublished results on the basis of high intensity of infection with T. colubriformis, with the exception of one farm (Meu). They were all located in the same area in west-central France and were within 50 km of each other, with the exception of one farm (Bau). The climate in the area is temperate with a yearly mean temperature 1 I°C, and over 600 mm mean rainfall. The main characteristics (either numerical or categorical) were derived from a questionnaire and an autumn visit to each farm. The most important are shown in Table 1. Numerical variables were: the age of farm, the number of goats bought at the establishment of the farm, the number of farms from which goats originated, the present number of dairy-goats, the percentage of alpine breed in the herd, the percentage of goats in their first-lactation, duration of kidding period, area of permanent and non-permanent pastures, and number of anthelmintic treatments per year. The number and eventual treatment of introduced goats or bucks in the last five years were also recorded. The categorical variables were: animal breeding is economically more important than agriculture, other animals than goats are bred in the farm, usage of a small pasture (less than 1 ha for 50 goats) intended for exercise, free access to such a pasture in winter. One or two goats from each farm were bought and necropsied in Autumn 1991. The procedures for necropsies and species identification, as well as the representativity of an individual goat were presented elsewhere (Cabaret and Gasnier, 1994). The percentage of T. colubriformis in helminthic fauna and intensity of infection were established on one or two goats (Table 1). The specimens of T. colubriformis from one goat from each farm were collected and preserved in liquid nitrogen until processed as described in N'Zobadila et al. (1993). Briefly, two polymorphic enzymatic systems were studied, malate dehydrogenase (MDH, E.C.1.1.1.37) and glucose-phosphate isomerase (GPI, E.C. 5.3.1.9), using thin-layer starch gel electrophoresis of individual male worms. Regression analyses between variables were performed with the least-square method. Breeding management or parasite population characteristics were transformed in pseudo-distances calculated as: absolute differences between values of each considered parameter in each possible pair of farms. These absolute differences were used to correlate breeding management and parasitic variability to genetic variability (Fst or genetic euclidean distances derived from principal component analysis, coded DEUCL). For example, genetic distance between the farms i and j was regressed to difference of size of herds in i and j farms. Multivariate analyses were usually preferred as variables were in most cases related to each other. The analyses used in this paper were principal component (description of farms based on numerical characteristics), discriminant (test of an a priori clustering of farms based on numerical characteristics) and correspondence analyses (description of farms based on categorical characteristics, classified as yes or

68 97 70 51 58 16 28 76 46 2

Bau Bet Cas Cuv Del Ech Gal Lel Men Meu

Low High High Medium High Medium Medium High Medium Medium

Intensity of infection a

7 110 50 70 38 100 50 30 60 18

No. of goats

0 0.02 0.16 0.04 0.01 0.14 0.17 0.20 0.10 0.14

Hectares of pasture per goat

a LOW (1-300); medium (301 - 1000); high ( > 1000 nematodes),

Percent T. colubrijbrmis in helminthic fauna

Dairy-goat farm

Table 1 Goat-farm breeding management and nematode infection

(n) n (y) n

(y) y

(n) n (y) n (Y) Y (y) n (Y) Y (n) n (n) n

Exercise yard (presence), access in winter 1 6 1 3.5 1 I 4 I 1 1

(early) (early) (early) (early) (late) (early) (late) (early) (late) (early)

Duration of kidding period in months (early/late) n n y n y n n n n n

Alternation of compounds for anthelmintic treatments

Animal breeding more important than agriculture

n

Other ruminants bred in farm

e~

L

e,

G. N'Zobadila et al. / Veterinary Parasitology 66 (1996) 213-223

217

no). The breeding management was analyzed by two-step multivariate analyses: combined variables obtained in principal component and correspondence analyses were used finally in discriminant analysis. Analyses were performed with a Stat-Itcf computer package. The significance of axes was derived from Lebart et al. (1977). The departure from Hardy-Weinberg equilibrium was estimated by exact Fisher test. The values of fixation indices of Wright (Fst between populations and Fis within population) were calculated using the method of Weir and Cockerham (1984) and their confidence interval using the jacknife resampling procedure (averages are calculated repeatedly on samples where one observation, chosen at random, is missing).

3. Results 3.1. Characteristics o f the f a r m s

The numerical farm variables were processed with principal component analysis, and the first three axes (i.e., the most efficient combination of variables) were used in further analyses. The same was done with categorical variables processed by means of correspondence analysis. A principal component analysis established on the combined variables (3 from numerical and 3 from categorical variables) did show that five groups of farms might be described. A discriminant analysis was then performed on these five groups of farms with the same combined variables. The first discriminant axis (significant at P < 0.05) was mostly derived from correspondence analysis combined variables CA1 and CA2; their equations were: CA1 = 0.56 E K + 0.21 A G - 1.71 OR - 2.43 E Y + 0.74 E W + 0.21 A T CA2 = 0.32 E K - 3.10 A G + 1.24 OR - 1.72 E Y - 0.56 EW + 0.21 A T

with, coded as 0 (absent) and 1 (present): E K = early kidding period (November to January). AG--animal-breeding economically more important than agriculture in the farm. OR = other ruminants than goats bred in the farm. E Y = existence of exercise yard. EW = access to exercise yard in winter. A T = alternation of different compounds for anthelmintic treatments. The discriminant axis is shown in Fig. la and could be described with variable CA1 on its left part and CA2 on its right one. The most important characteristics were: the importance of animal breeding versus agriculture; the presence of other ruminants than goats; free access to an exercise yard in winter. 3.2. Genetic study

Three distinct phenotypes could be observed in MDH: single banded (allozymes of rapid or slow migration) and triple-banded patterns, characteristic of heterozygotes for dimeric enzymes. The quaternary structure of GPI could not be accurately assessed, but individuals with intermediate bands located from slow to rapid migrating individuals were scored as heterozygotes. The allele frequencies for the two putative loci in the populations harboured by the goats in 10 farms are given in Table 2. These frequencies

218

G. N'Zobadila et al. / Veterinary Parasitology 66 (1996) 213-223

Table 2 Genotypes (RR, SS and heterozygote RS) and allelic frequencies (Rapid, R and Slow, S) for glucose-phosphate isomerase (GPI) and malate dehydrogenase (MDH) in populations of Trichostrongylus colubriformis from 10 dairy-goat farms (coded with three letters) Enzyme

GPI Ban Ber Cas Cuv Del Ech Gal Lel Men Meu MDH Bau Ber Cas Cuv Del Ech Gal Lel Men Meu a

Genotype

n

RR

RS

SS

14 13 10 13 7 10 9 5 6 5

3 4 6 6 9 3 18 13 13 15

15 3 5 1 8 7 0 1 7 12

17 10 21 15 27 12 38 16 37 19

4 4 6 3 3 7 3 5 1 6

1 6 4 4 1 2 0 2 0 1

Allelic frequency

Departure from Hardy-Weinberg (Fisher test) a

R

S

32 20 21 20 24 20 27 19 26 32

0.67 0.75 0.62 0.80 0.48 0.58 0.67 0.61 0.48 0.39

0.33 0.25 0.38 0.20 0.52 0.42 0.33 0.39 0.52 0.61

s ns ns ns ns s s ns ns ns

22 20 31 22 31 21 41 23 38 26

0.86 0.60 0.77 0.75 0.92 0.74 0.96 0.80 0.99 0.85

0.14 0.40 0.23 0.25 0.08 0.26 0.04 0.20 0.01 0.15

ns s s s ns ns ns ns ns ns

ns, not significant; s, significant.

varied considerably b e t w e e n farms: 0 . 3 9 - 0 . 8 0 for GPI rapid allele and 0 . 6 0 - 0 . 9 9 for M D H rapid allele. Departure f r o m the H a r d y - W e i n b e r g equilibrium (Fisher exact-test) was recorded on a few occasions. The analysis o f genetic differentiation within and b e t w e e n populations was calculated by W r i g h t statistics, respectively F~s and F~t. The jacknife estimate o f m e a n Fst was 0.06. Fis varied f r o m - 0.49 (Gal) to 0.82 (Bau) for GPI and f r o m - 0 . 0 3 (Gal) to 0.65 (Cuv) for M D H . This m e a n t that the farms were apparently different f r o m each other (Fst ~ 0) and that small to large departures f r o m H a r d y - W e i n b e r g were recorded, due to a deficiency in h e t e r o z y g o t e s (Fis > 0) or excess in h e t e r o z y g o t e s (F~s < 0). As the farms were different on allele f r e q u e n c i e s and by heterozygote frequencies, a principal c o m p o n e n t analysis was p e r f o r m e d on the genotypic frequencies (Fig. lb). The first axis (i.e., c o m b i n a t i o n o f g e n o t y p i c frequencies constructed in such a way that m a x i m u m variance could be obtained) represented 59% of variance; due to small size o f data matrix, it was the only significant axis at P < 0.05. The farms located on the left part o f the axis were characterized by high f r e q u e n c y o f h o m o z y g o t e o f slow M D H and high frequency o f rapid GPI. The farms on the right part o f the axis were characterized by high f r e q u e n c y o f h o m o z y g o t e rapid M D H and h e t e r o z y g o t e GPI.

G. N'Zobadila et al./ Veterinary Parasitology 66 (1996) 213-223 a

I Breeding management

219

}

(Discriminant analysis) Cas Ech

Cuv Lel

Meu Bau

Men Gal

0 r • Animal breeding more Important ~ than agriculture • Other animals than goats • Free access to exercise yard ~ In winter

Del

Axls 1 _-(100% Less importen

'1

Only goats | No free access|

J

{ AIIozymes ]

b

(Principal component analysis) Ber Ech Cuv

II

Ii

Cas

Bau Lel Del

, Itl 0

I

Gnl Men

l l'-

Meu

Axis 1

(59%) MDH s l o w GPI rapid

==

MDH rapid GPI heterozygote

Fig. 1. (a) Breeding management: similarities between farms (abbreviated to three letters) based on discriminant analysis. (b) Allozyme frequencies of Trichostrongylus colubriformis (MDH and GPI): similarities between farms (abbreviated to three letters) based on principal component analysis.

3.3. Breeding management, parasitism and genetic variability The two estimators of the genetic variability estimated from Fst or DEUCL were highly correlated ( r = 0.51): Fig. 2. The genetic differences between farms as assessed by means of Fst were not correlated to kilometric distances between farms, nor to the differences between percentages of T. colubriformis in the worm population, numbers of hosts in herds, areas of permanent pasture available to goats or durations of kidding period as shown in Fig. 3. Similar findings were obtained with euclidean distances (DEUCL: distance between farm i and j) derived from principal component analysis of genetic data. However, highly significant correlation was recorded between DEUCL and the difference in size of goat herds (NGOATS: absolute value of No. of goats in herd i minus No. of goats in herd j):

DEUCL = 2.95 + 0.129 NGOATS r 2 = 0.42; P = 0.001. Similarities were easily found between the ranking of farms based on breeding management or allozymes (Fig. lb). The farms Cas, Ech and Bet were located on the extreme of left side and the Gal and Men on the right side of first axis for each characterization (breeding management Fig. la or genotypic frequencies Fig. lb). The

220

G. N'Zobadila et al./ Veterinary Parasitology 66 (1996) 213-223 0.3

r=0.51

o

0.2

o

o

co LL

e

0.1

o o

0.0

o

-0,1 0

o o

~

o

¢,

o

e

r 5

p 10

15

20

DEU©L Fig. 2. Significant linear relationship between genetic diversity estimates. Wright Fst and Euclidean distances obtained from discriminant analysis (DEUCL) in ten goat dairy-farms.

euclidean distances between farms were established by pairs (46 pairs), either on analyses performed on breeding management or allelic frequencies, in order to test the similarity of both rankings. The correlation coefficient (0.70) between distances obtained

DEI.~L

FST °.

KM





o•°

O

O

o



,



•"

.o ~

.•



~.o

°

t

°

°

COL

...;:_'.:.



NGOAT8

HAGOAT



.--~..". :~.. °

-.

DORKID

Fig. 3. Relations between genetic differences (estimated by means of Wright Fst) and distance in km between farms, % of Trichostrongylus colubriJbrmis in worm community = col, and several characteristics of breeding management. (No. of goats = ngoats; no. of ha of permanent pasture per goat = hagoat; duration of kidding period = durkid).

G. N'Zobadila et a l . / Veterinary Parasitology 66 (1996) 213-223

221

from breeding management (Euclidean distances derived from discriminant analysis) and allelic frequencies (DEUCL) was highly significant (P < 0.00).

4. Discussion

4.1. The specific unity of T. colubriformis Heterozygote deficiency was recorded in several farms but it was not of the same magnitude as previously described by N'Zobadila et al. (1993) on a several farm pooled sample. In the latter case, the high level of heterozygote deficiency was certainly due to the Wahlund effect (false heterozygote deficiency due to pooling of populations having different allelic frequencies). The deficiencies in heterozygotes found in several farms are difficult to explain. The existence of cryptic species could explain this deficiency of heterozygotes but remains highly speculative as they were not consistent from one loci to another. The recorded deficiencies/or excess might have been generated by hazard of sampling. The most probable hypothesis is that T. colubriformis is one single species, that may show genotypic and phenotypic variabilities according to selective pressures.

4.2. The reliability of a single goat necropsy to represent farm genetic nematode variability The within-flock genetic variability is very limited in another trichostrongylid nematode Teladorsagia circumcincta, also a parasite of sheep and goats (Gasnier et al., 1996): the Fst values are less than 0.010 in between-sheep populations. This was corroborated by the fact that allozyme frequencies (GPI, MDH and three other enzymes) values from these nematodes collected from one host to another, in the same flock, were very similar (Gasnier et al., 1996). This could indicate that the larval sampling on the pasture is equally well-done by the different lambs. The efficient role of ewes as larval sampler on a pasture has been demonstrated previously (Cabaret et al., 1986). The result of such an efficient sampling has also been shown on species diversity encountered in dairy-goats (Cabaret and Gasnier, 1994). We do not mean that one goat represents the intensity of infection of the herd, but that it may well represent the assemblage of species or the genetic variability of a species of nematode.

4.3. Relationship between T. colubriformis high frequency in a farm strongyle community and genetic variability It could be hypothesised that the farms with high proportion of T. colubriformis are favourable to that species, so that within population genetic diversity is large, whereas in less favourable conditions the genetic diversity is reduced. Although a tendency in that line could be observed in Fig. 2, it was not statistically significant. The range of T. colubriformis frequency in the community in our selected farms is limited if we exclude the Meu farm: 16-97%, and 7 farms out of ten had figures over 45%. The hypothesis that low species richness (high prevalence of one species, in that case T. colubriformis)

222

G. N'Zobadila et al./Veterinary Parasitology 66 (1996) 213-223

results in high within dominant-species genetic variability, could not be demonstrated in our samples but deserves further investigations in a larger array of specific diversity.

4.4. Relationship between genetic and breeding management variabilities The allozyme frequencies of T. colubriformis were different among the investigated farms. They were established on only two polymorphic enzymes and their variations should be considered only as an attempt to assess genetic variability. The farms could be tentatively ranked on these allozyme frequencies and we may then wonder if this ranking was the result of stochastic fluctuations or corresponded to selective pressures in relation to the characteristics of farms. The ranking of farms by means of Wright Fst or euclidean distances derived from principal component analysis was highly correlated (Fig. 2). The latter was apparently better related to breeding management traits (see Fig. 3). The fact that genetic variability is related to breeding management indicates that the chosen allozymes are not neutral (did not fluctuate randomly) but respond to selective pressure in the herd. A similar response to selective pressure was recorded in another strongyle, Teladorsagia circumcincta, both in laboratory reared and in wild populations (Gasnier et al., 1992) The fact that multivariate ranking of farms by means of breeding management was apparently more efficient than separate bivariate analyses in assessing the relationship between management descriptors and genetic ones, indicates that breeding management characters were interrelated. Two of the discriminating management traits were very general (economical importance of animal breeding and presence of other ruminants) and did not support any clear biological hypothesis. Conversely the genetic diversity (DEUCL) was related to diversity in host number in each farm and could possibly be better interpreted in terms of biological hypotheses. The number of goats might be a possible clue to explain diversity: larger number of goats corresponds to larger natural pastures and to more numerous origins of goats that were bought at the establishment of the farm (Cabaret and Gasnier, 1994). These characteristics are possibly related to larger genetic diversity that should be investigated by a larger array of isoenzymes or more polymorphic markers, such as microsatellites. The fact that isolation by distance (DEUCL or F~, in relation to distance in km between farms) was not evident is probably due to the helminthic isolation of farms for long periods. This isolation was real as far as domestic hosts were concerned; the role of lagomorphs in the transportation of T. colubriformis from one farm to another remains probably of limited importance although experimental infections reported high infection success (Ciordia et al., 1966). The absence of relation between distance and genetic diversity between farms (see Fig. 2) tend to prove that farms do not exchange T. colubriformis. The farms located in the same area should present genetically similar helminths, whereas those very distant should have limited opportunity to exchange helminths, and present high genetic diversity, which was not the case. We may then propose the following scheme: at the establishment of the farm, goats from various origins are introduced and the variability is maintained or reduced depending only on breeding management, a large herd being a guarantee of genetic diversity maintenance.

G. N'Zobadila et al. / Veterinary Parasitology 66 (1996) 213-223

223

Acknowledgements Financial support in the form of Ph.D. grants from the French Ministry of Research (N.G.) and Congo state (G.N.) is gratefully acknowledged. The Central Region of France partly funded this Research. We thank C. Sauv~ and J. Cortet (INRA, Nouzilly) for technical help and C. Chartier (Station de Pathologie Caprine, Niort) for providing information on the dairy-goat farms in the Poitou-Charentes region of France.

References Andrews, R.H. and Beveridge, I., 1990. Apparent absence of genetic differences among species of Teladorsagia (Nematoda: Trichostrongyloidae). J. Helminthol., 64: 290-294. Bryant, E.H., 1974. On the adaptative significance of enzyme polymorphisms in relation to environmental variability. Am. Naturalist, 108: 1-19. Cabaret, J., Anjorand, N. and Leclerc, C., 1989. Parasitic risk factors on pastures of french dairy-goat farms. Small Ruminant Res., 2: 69-78. Cabaret, J. and Gasnier, N., 1994. Farm history and breeding history influences on the intensity and specific diversity of Nematode infection of dairy-goats. Vet. Parasitol., 53: 219-232. Cabaret, J., Mangeon, N. and Graner, L., 1986. The larval uptake of digestive strongyles by sheep on pasture: estimation by the mean of oesophagai fistulae. Vet. Parasitol., 19: 315-320. Cain, A.J. and Provine, W.B., 1991. Gene and ecology in history. In: R.J. Berry, T.J. Crawford and G.M. Hewitt (Editors), Genes in Ecology. Blackwell, Oxford, pp. 3-28. Ciordia, H., Bizzell, W.E., Porter, D.A. and Dixon, C.F., 1966. The effect of culture temperature and age on the infectivity of the larvae of Trichostrongylus axei and T. colubriformis in rabbits and guinea pigs. J. Parasitol., 52: 866-870. Gasnier, N., Cabaret, J. and Moulia, C., 1992. Allozyme variation between laboratory reared and wild populations of Teladorsagia circumcincta. Int. J. Parasitol., 22: 581-587. Gasnier, N, Cabaret J., Sauv~ C. and Gruner, L., 1996. Host, season and year do not play an important role on genetic variability in the trichostrongyle nematode Teladorsagia circumcincta as assessed from ailozymes. Comptes Rendus de l'Acad6mie des Sciences, Paris, D, 319:113-118. Lebart, L., Morineau, A. and Tabard, N., 1977. Techniques de la description statistique. M6thodes et Iogiciels pour I'anaiyse des grands tableaux. Bordas, Pads. Nadler, S.A., 1990. Molecular approaches to studying helminth population genetics and phylogeny. Int. J. Parasitoi., 20:11-29. N'Zobadila, G., Cabaret, J. and Gasnier, N., 1993. Allozyme variations between sheep or rabbit laboratory reared and goat wild populations of Trichostrongylus colubriformis. Int. J. Parasitol., 23: 1087-1089. Powell, J.R., 1975. Protein variation in natural populations of animal. Evolutionary Biol., 8: 79-119. Weir, B.S, and Cockerham, C.C., 1984. Estimating F-statistics for the analysis of population structure. Evolution, 38: 1358-1370.