A case of parasite-mediated competition? Phenotypic differentiation among hookworms Uncinaria sp. (Nematoda: Ancylostomatidae) in sympatric and allopatric populations of South American sea lions Otaria byronia, and fur seals Arctocephalus australis (Carnivora: Otariidae)

Marine Biology 112,527-533

(1992)

©

Springer-Verlag 1992

A case of parasite-Inediated cOlnpetition? Phenotypic differentiation among hookworms Ullcillaria sp. (Nema toda: Ancylostomatidae) in sympatric and allopatric populations of South American sea Iions Otaria byrollia, and fur seals Arctocephallls allstralis (Carnivora: Otariidae) M. Gcorgc-Nascimcnto

1, IVI.

Lima

2

and E. Ortiz

3

I Departamento de Ciencias del Mar (BIOTECMAR), Pontilieia Universidad Calóliea de Chile, Sede Taleahuano, Casilla 127, Talcahuano, Chile 2 Industria Lobera y Pesquera del Estado de Uruguay (1LPE), Rambla Ballasar Brum s/n esq. Feo. Tajes, Monlevideo, Uruguay 3 Fisheries [{eseareh [nstilule, 202 Nieholson Pole, P.O. Box 2\, Cronulla, New South Wales, Australia 2230 Date of fínalmanuseript

aeeeptanee:

September

11, 1991. Communieated

Abstract. Hookworms in otariids are considered to meet ¡he conditions and to fulfill the predictions set forth in the theoreticalliterature to define a likely case of host coexistence mediation by a shared pathogen. The intensity of infections, the prevalence al' skin lesions and the morphometry al' hookworms Uncinariaspp. were examincd in South American sea lions and fuI' seals sampled along the Chilean and the Uruguayan coasts in spring and summer 1981 to 1991. [n sYl11patric host populations from Uruguay, there were clear differences in intensity al' infections, prevalence of skin lesions, and body size al' the hookworms from the twa host speeies. Sea lions from Chile, allopatric to fuI' seal populations, are less intensively inf'ccted and hookworl11s f()und in these sca lions have the smallest body size reported in otariids, while those from South Amcrican fuI' seals are the largest. Hookworms found in sea lion pups from populations sympatric to fur seals revealed intermediate values in intensity and in body size, and the sea lion pups had tbe higbest prevalence 01' skin lesions. Other reports of hookworms in otariids show intermediate morphometry, fol. lowing a gcncral linear trend 01' dilTerentiation in size. Consequently, they are considered to belong to the same, widely distribllted species Uncinaria lllcasi Stiles, 1901,

Introduction Parasites would affect the olltcome of the coexistence of closely related host species if they have a differential impact 011 their demographic rates (Holt and Pickering 1985). This olltcome mimicks the results of exploitative competition 1'01'limited rcsourccs, and wOllld favour host species with wider geographic range (the Host Geographic Range hypothesis), and also species smaller in body mass (the Host Body Size hypothesis, Price et al. 1988). This situation may also be conditioned by the type of life-cycle the parasite exhibits (indirect, direct), and their mode al' transmission (Holt and Pickering 1985), Howev-

by O. Kinne, Oldendorf/Luhe

el', it is first necessary to demonstrate that, at least using traditianal classification criteria, the parasites collected in different host species belong to the same species. The analysis should allow for the possibility that when a parasite is exposed to different selective pressures in the alternative host environments phenotypic as well as genotypic differentiation may result (Holmes 1976). Hookworms are nematodes which are parasitic during the adult stage in the intestine 01' mammals. They have a direct life-cycle including free-living (eggs and larvae) as well as parasitic stages (tissue-dwelling larvae and intestine-dwelling adults, Olsen 1974). Among marine mammals, olariids are hosts 01' at least two hookwoorm specics: Uncinaria lucasi Stiles, 1901, rccorded only in the northern fur seal Callor/¡inus ursinus and the Steller sea lion ElIlI1etopias jllbata (Olsen 1958), and Uncillaria /¡all1iltolli Baylis 1933, recorded mainly in sea lions from a variety 01' host species and geographical localities: the Steller sea lion and the Ca!ifornian sea !ion Zalop/¡lls californialllls in California (Bay!is 1933, 1947, Dailey and Hill 1970), the South American sea lion Otaria byrollia (O . .!lal'csccns Shaw, see Oliva 1988), in the Falkland (Malvinas) Islands (Baylis 1933), in Uruguay (Botto and Mañé-Garzón 1975), and the Australian sea lion Neop/¡oca C'Ínerea (Beveridge 1980). According to Baylis (1933) and Dailey (1975), there could still be at least two ather undescribed Ullci/lária "'species found in phocid host5: one in the southern sea elephant Mirollllga leonilla, and another in the ringed seal Pusa /¡ispida. Bionomie studies have only been canied out 011 U. lllcasi. The lifecycle of this species is synchronized in reproduction to their h05ts, being transmitted to just-born northern fur seal pups by their mothers, mainly during early lactation (vertical transmission). Free-living infective larvae are also able lo actively penetra te the hosl skin and migrate to the intestine through the host tissues, thus constituting a horizontal mode of transmission, and the only way interspecific transmission may occur (Olsen 1958, 1974, Olsen and Lyons 1965, Lyons and Keyes 1978, 1984, Lyons 1983).

:;

M. George-Nascimento

528

Existing records show that hookworms are able to live in sympatric host species (Olsen 1958, Olsen and Lyons 1965), as well as in the same or distinct host species in differcnt gcographical localities. The hookworms found in a rangc 01'spccics ami dilTcrcnt localitics reveal diverse degrees 01' l11orphological differentiation (Botto and Mañé-Garzón 1975). However, despite the variety 01' host and geographical records, thorough investigations 01' the 1110rphological variability 01' hookworms in pinnipeds are rare because the numbers 01' parasites examined are usually low or not indicated. Although morphological variability and its potential causes have not been explicitly ql1antified, one report on the variability 01' hookworl11 body size 01' tissue-dwelling third-stage larvae, revealed that host populations categories (pups, males, etc.), as well as host species, may be important in explaining its variance (Lyons and Keyes 1978). In this paper we speculate on the possibility that Uncinaria mediate host coexistence, according to the conditions posed by Holt and Pickering (1985), and we consider the extent to which our observations fit or deviate from the predictions posed by Price et al. (1988). We investigate the hypothesis that the hookworms reported in pinnipeds correspond to one widely distributed species, U. ¡ucasi, with intraspecific morphological variations due to gcographical as well as lo hosl infll1enccs. We compare mor¡)hometric features 01' adult hookworms reported in the literature with those collected in South American sea lions fram populations sympatric to South American fur seals (Arctoceplzalus australis) in Uruguay, and also to hookworms collected in sea lions from central Chile, in populations allopatric to fur seals. Finally, we assess the statistical significance 01' the difl'erences in hookworm abl1ndance, and in prevalence 01' skin lesions, potentially atlribl1lable to hookworm infeclions, belween pups 01' the two host species.

et al.: Parasite-mediated

competition

Fig. l. Map showing study siles in the Bio-Bio Region in cenlral Chile (Ieft), and the Isla de Lobos, Uruguay (righl)

Materials and methods Sampling siles are lhe largesl colonies in lhe sludy areas and represenllhc silualions sclcclcd 1'01' conlrasl: places where sea lions are . sympalric or allopalric lo fur scals. In Cobquccura, lhere are ca. , 5000 sea lions, ami no fur seals have been reporled (George-Nascimenlo el al. 1985). On lhe conlrary, Isla de Lobos is the largesl mixed colony in lhe Atlantic Ocean, wilh ca. 24000 fur seal pups and 2 500 sea lion pups bcing reared annual!y (M. Lima personal observalion). During spring and summer 1981 and 1982 we sampled 30 adull sea lions from centra] Chile. In summer 1982, during the sea lion breeding season in cenlral Chile, nine just dead 01' evidently weak sea lion pups were sampled in Cobquecura for hookworms (Fig. 1). The seleclion ofweak pups was intended to improve the probability 01' finding parasites. In summer 1988, we sampled three more sea lion pups and in winler 01' the same year, nine adult males, al! from Isla de Lobos, Uruguay. Finally, in January 1991 we sampled an addilional six sea lion pups al Uruguay, and nine more al Cobquecura (Fig. 1). Twelve adull male fur seals amllhree pups were sampled in Uruguay during summer 1988. The lalter came from an area not exceeding 100 m2 Sea lion pups were also sampled from the same area at lhe same time. During winler 1988, we additionally sampled 26 adull males and six fur seal yearlings. In January 1991 we sampled an addilional 12 fur seal pups. In al!, we examined 66 sea lions and 59 fur seals (Table 1).

r

The digeslive lracl 01' eaeh host individual was opened wilh an incision from the oesophagus to the anus. The contents 01' lhe traet were washed lhrough a .33-mm sieve using tap water, and the malerial retained by the sieve was examined under a stereoscopic microscope for parasiles. Parasites were fixed in hol 5% formalin, and hookworms were examined under a microscope arter clearing in lactophenol. Infrapopulation size was compared between samples from Chile and Uruguay by means 01' a Kruskal- Wallis lesl carried oul across hosl species. Vertical lransmission 01' hookworms is exclusively reslricled wilhin hosl species. Hence, lransmission between hosl species must be due to skin penetration by frec-living larvae, that is, horizontal lransmission. Consequently, in order lo compare relalive rales 01' interspecific transmission, 4100 pups 01' bolh species (4000 01' which were fur seals), were examined 1'01' lesions in the ventral skin during a pup-marking program in summer 1989 at Isla de Lobos. The proporlion ofpups ofboth host species in the samples accurateIy represent their relative abun.dances in lhe island . Elevenmorphomelric variables were rccorcled in 93 worms. The following five morphomelric rneasuremenls were recorded 1'01' bolh sexes: body lenglh, body width, oesophagus length, lenglh and width ofbuccal capsule. For male parasites lhe fol!owing addilional morrhomelric variables consliluling sexual sccondary characters were also measured: spicule lenglh, and the ralio belween the length 01' lhe bifurcalion 01' the dorsal ray in the copulalory bursa to its totallength. The width and lenglh 01' eggs, distance from vulva to poslerior end. and laillenglh, were recordedfor females. Measurements 01' hookworms from Chile were nol recorded for each individual worm separately. The samplesfrom Uruguay were complele and amenable to mullivariate analyses ofmorphometric measurements. Mullivariale analysis was used lo delecl the polenlial alloxenicsympalric effecl, that ¡s, lhe effcct on parasite morphology due to differenl host species in lhe same gcographic localily. In arder lo delerminc which variables bcsl combine to explain lhe morphomelríc patterns when projecled on a reduced space, we employed the five morphomelric variables common to bolh sexes lo perform a principal component analysis (PCA).on lhe correlalion matrix calculalcd with lhc data from fur seal and sea lion pups collecled in Uruguay. We kcpl orthogonalily cif lhe axes by employing Varimax rotation (Legendre and Legendre 1983). In this PCA we excluded

M. George-Naseim"ento

et al.: Parasite-mediated

529

eompetition

observations by random elimination, in order to make them more similar in sample size for eaeh sex (nine males and 12 females per host speeies), lhus rendering the dala more suitable for analysis. Two other PCAs were run for eaeh sex separately. The data for these analyses included lhe l11orphomelrie variables exc\usive to eaeh sex and alllhe obscrvalions availablc (30 male and 30 female from fur seals, and nine malc am! 12 female from sea lions). Thc synxcnie-allopatrie cffcel, thc cffcet on para site morphology allribuled lo differenl gcographie populalions 01' the same host speeies, was assessed by means 01' eomparing our measuremcnts on lhe morphomelry 01' hookworms in Olaria hyronia from Chilc and Uruguay wilh rccords for the samc host spccies in thc literature. We al so assessed the alloxenie-allopatrie effcct, theeffeet related to dilTercn( host specics and geographic loealitics, by eomparing our data \Vith those for olhcr host specics and geographieallocalities. We did lhis by f'irsl labulaling lhe basie descriptivc statistics (arithmetic means, ranges, and ratios \Vilh respeet to body length) for the previously mentioned variables. We thcn seleeted the best single regression model from the matrices of correlation between the morphomelric variables in our dala, and supcrimposcd it onto a plot with previous rccords Of hookworms in pinnipeds. A similar approaeh \Vas employed by Oailcy (1975) for lhe analysis 01' intraspecifie variation 01' lungworms in pinnipeds (Parafi/aroides spp.).

ResuIts One adult female and five out of 18 sea lion pups were found with hookworms in Chile. In Uruguay, two out of nine adult males, and four out of nine pups sampled in the summer were infected. Of the fur seals, one out of 26 adult males sampled in summer 1988 harbored hookworms. All three pups sampled in summer of the same year, and ten out of 12 pups sampled in summer 1991 were infected. None of the six yearling fur seals sampled in winter 1988 at Uruguay were infected (Table 1). These data show no significant dilTerences in the prevalence of hookworm infections belween fur seal and sea lion pups caught in summer in Uruguay, but intensity of infections was most severe among fur seals sampled in summer, and leasl severe among sea lions from Cobquecura (Fig. 2, Kruskal-Wallis test, Chi-square ca. 17.00, df=2, P