Low Genetic Variability in the Hawaiian Monk Seal

Low Genetic Variability in the Hawaiian Monk Seal MARIA B. KRETZMANN,*+ WILLIAM G. GILMARTIN,$ AXEL MEYERJ GERARD P. ZEGERS,* STEVEN R. FAIN,$ BRUCE F. TAYLOR,$ AND DANIEL P. COSTA* Biology Board. University of California, Santa Cruz, CA 95064, (.".S.A. TDepartment of Ecology and Evolution. State Uni~ersity of New York, Stony Brook, NY 11794, U.S.A. #Hawaii Wildlife Fund, P.O. Box 540, Volcano, HI 9<n785. U.S.A. Â§U,S Fish and Wildlife Service Forensics Laboratory, 1490 East Main Street, Ashland, OR 97520. L'.S.A

Abstract: The Hawaiian monk seal (^Mondchus schauinslandi) is a critically endangered species that has failed to recovwjrom human exploitation despite decades ofprotection and ongoing nzanagement efforts de- signed to increase population growth. The seals breed atfive principal locations in the northwestern Haivai- law islands, and inter-island migration is limited. Genetic variation in this species is expected to be low due to a recent pop111c~tion bottleneck and probable inbreeding within small subpopulations. To test the hypothesis that small population size and strong site fidelity has led to loir within-island genetic variability and significant between-island differentiation, use used two independent approaches to quantify genetic variation both within and among the principal subpopulations Mitochondria1 control region and tRNA gene sequences (359 base pairs) were obtained from 50 seals and revealed very low genetic diversity (0.6% variable sites), u'ith no evidence of subpopulation differentiation. Multilocus DYA fingerprints from 22 individuals also indicated low genetic variation in at least some snbpopulations (band-sharing values fur ";t?zrelatecl" seals /ram the same island ranged from 49 to 73%), This method also providedpreliminary evidence ofpopulation subdivision (ft estimates o f 0.20 and 0.13 for two adjacent island pairs). Translocatio~~s of seals among islands may therefore have the potential to relieve local inbreeding andpossibly to reduce the total amount of variation preserved in the population. Genetic variation is only one of many factors that determine the ability o f an endangered species to recover. Maintenance of existing genetic diversity, however, remains an important prior- ity for conservation programs because of the possibility of increased disease resistance in more variable populations and the chance that inbreeding depression may only be manifest under adverse environmental conditions.

Baja Variabilidad Genetica en la Foca Hawdiana

Resumen: La foca b a ~ ~ ~ a i a n a es una especie enpeligro de extindon cuyapoblaci6n nunca se ha restableddo completamente cle sit explotacion hunzuna a pesar de decadas de proteccidn y 10s sucesiros programas de gesti6n disei3adospara a m e n t a r el crecinziento de lapoblacih. La crianza de /as focas occure en cinco islas del archipielago hauraiano, y la migration entre las islas es lirnitcida, Se esperaba que la variacih en esta especie fuera baja d e m o a tin reciente cuello de botella y a1 e~~trecruzunziento probable entre las subpobla- clones pequenas. Para probar la hwtesis de que el tunza@o pequdo de la poblaci(jiz y el fuerte arraigo a su Sugar natal ha ocasionado baja variabiliclad genetica dentro de las islas, y una diferenciacion sigmficativa entre las islas, usanzos dos nzdtodos independientes para medir la variabilidad genktica dentro de y entine las subpoblaciones prindpales. Se obtuvieron secuencias de la region de control del A m r i~zitocondri~il (359 pares de buses) de 50 focas, las cuales rerelaron una diversidad gendtica muy baja (0.6% posiciones rari- ablesj, sin euzdencia de uno diferenciacion subpoblaciotzal. "Huellas c/igilales" de ADN cle 22 indivuiuos tam- h im mos twon baja variabilidad gendtica, por lo menos en ciertas subpoblaciones (la pnipoim'in de flag- mentos combartidos bor focas "sin Darentesco" de la misma Ma vari6 entre el 49 v el 7 W . Este metodo

Address correspondence to Gloria Kretnnufln, Department of Ecolog)' and Evolution, State University o f ik-eif,' York, Stony Hrook. X Y 11 3 4 ,

l<.Y.A., email [email protected]~ysb.edu Paper submitted February 6. 1996 revised manuscript accepted June 27, 7996.

Conservation Biology, Pages 482-490 Volume 11, No, 2, April 199-

Krelzmunn el at. l.oir Genetic Vnridhititj' in Monk Seals 48.3

Introduction

The importance of genetic variation for predicting the vulnerability of an endangered species to extinction is a controversial issue in conservation biology (e.g., Laiicle 1988; Car0 & Laiirensoii 1994; Merola 1994; O'Brien 1994). Inbreeding and the resulting loss of genetic variability arc often inevitable consequences of the severe reductions in population size experienced by endan- gcrecl species, but the extent to which this results in inbreeding depression is likely to vary tremendously among species and is difficult to document in wild pop- illations. Even more difficult to assess is the importance of the loss of genetic variation for the "adaptive potential" and long-term prospects of populations. Neverthe- less, a primary goal of conservation efforts has been to preserve a maximum amount of genetic variation in each species under consideration, based on the per- ceivccl link between a lack of genetic variation and an increased extinction risk for small populations (e.g., Bearelmore 198.3; Gilpin & SonlC 1986; Vrijcnhock 1994; Frankham 1995).

I'he Hawaiian monk seal (Monuchus scbuui~s~utldi) is an endangered species that currently numbers approximately 1.300 inclivicliitils (W.G. Giln~tirtin, unpub- lished data). This species was hunted to near extinction in the nineteenth century (Kenyon & Rice 1959), hut the size and duration of the population bottleneck are not well documented. Following partial recovery, an additional decline of about 50(%, (from approximately 5000 to 1500 individuals; Kenyon 1972; Johnson et al. 1982; Ragen 1993) occurred between tlic late 1950s and tlic 1970s, which lias been at least partly attributed to liu- man disturbance at breeding rookeries (Kenyon 1972; Gerroclctte & Giln~artin 1990). ' 1 ' 1 ~ seals breed at five main locations in the northwestern Hawaiian Islands (Fig. I), and they show a high degree of site fidelity (Johnson & Kridler 1983). Although about 10% of adult monk seals have been sighted at a location other than tlicir natal island (Ragen 1995), the degree to which these migrants may contribute to the local gene pool (i.e., tlieir reproductive success) is unknown. Given tlieir history of small population size and relatively isolated subpopulations, monk seals are expected to be relatively low in genetic variability clue to the effects of ran- dom genetic drift and inbreeding.

The combination of sn~all subpopultitioiis and low inter-islancl migration rates is also likely to lead to significant genetic differentiation among subpopulatioiis, thereby preserving more genetic variation overall than would exist in a panmictic population (Allendorf 1985; Varvio et ill. 1986; Lacy 1987). The importance of a low migration rate among subpopultitions in order to ensure the preservation of alleles has been emphasized in rec- onimenclations for endangered species management because hcterozygosity can he reconstituted by artificially increasing gene flow if inbreeding depression is a scri- oils concern (e.g., Allenclorf 1983; l.acy 1987). Recently, however, monk seal population managers have effec- tively raised migration rates by translocating seals among islands, primarily in order to relocate starving juveniles from French Frigate Shoals to locations where food avail- ability may he greater and survival better (Gerroclette &

Giln~artiii 1990; Gilmartin & Eberliardt, 1995). Although these relocations have increased reproductive potential for the population (Giln~artin & Eberharclt 1995), the ag- gregate potential effects of these nianagement actions cannot be properly evaluated in the absence of data on the genetic Â¥itructur o f the monk 5cal population

Figure 1. M a p ofthe northwestern Hawaiian islands; arrows indicate islands inhabited by movk seals. Sam- pling u'cls conducted at the five sites currently used for breeding (b'refich Frigute Shoals, L L I J ~ I ~ I , Lisiat~ski, Pearl and Hermes Reef; and Kure Atoll). Dotted lines around islands represent the 100~fa iso/>uth. Adapted from Kugen ( I 99-31 with permission.

To test the hypothesis that small population size ancl strong site fidelity has led to low within-ishind genetic variability and significant between-island differentiation in the Hawaiian tnonk seal, we used two independent approaches to quantify genetic variation both withiii and among the principal monk seal siihpopulations. Both techniques involve selectively neutral genetic markers; we assume tliat these provide an index of gc- notne-wide variation, witli possible significance for the survival potential of tlie population. We examined the sequence of the non-coding mitoclionclrial DNA (mtDNA) control region, the fastest evolving portion of the mito- chonclrial genome and often a sensitive gauge ol' intraspecific genetic variation (e.g., Koclicr et al. 1989; Mcyer 1994). Because many other studies involve sequencing the same segment of DNA, these data also provide a "universal metric," thereby facilitating the interpretation of tnonk seal genetic variability in tlic context of what is known about oilier similar species. Such a comparative approach, including species which have never been threatened witli extinction, and those which are recov- ering to varying degrees, may shed some light on the importance of genetic variation in the conservation of en- dangcrcd species.

Several authors (e.g., Cronin 1993; Moritz 1994) have stressed tlie in~portance of using nuclear DNA markers in conjunction with tntDNA analysis in conservation ap- plications because of the varying niodes of inheritance and population dynamics of the two types of genetic ma- terial. Maternally-inherited n1tDNA is more prone to genetic drift (and therefore more sensitive to population bottlenecks) than arc nuclear loci due to an effective population size one-quarter that of nuclear DNA (Birky et al. 1989). Accordingly, we used a niultilocus DNA fingcr- printing approach, which is based on tandenily repeated nuclear DNA sequences known as niinistitellites. Most minisatellites arc highly polymorphic clue to variation in copy number of the repeat unit, and a single probe based on a shared "core" sequence can detect many variable loci simultaneously (Jeffreys et al. 1985). Although this approach provides a level of resolution often best suited to identify individuals and close relatives, in some species witli relatively low genetic variability DNA fingerprinting has been used successfully to examine relatioii- ships among siibpopiilations (Gilbert ct al. 1990; Triggs ct al. 1992).

Sampling

Tissue samples were collected from the rear flippers ot seals during tagging operations at the five principal breeding sites of the Hawaiian monk seal (French Frigate Shoals, Ltiysan Island, Lisianski Island, Pearl and Hermes Reef, and Kure Atoll; see Fig. 1) and were frolen as soon

as possible, although sample preservation was not en- tirely successful at these remote field sites. Almost all samples were from pups of tlic year; the exceptions were two juveniles (from Lisianski, Pearl and Hermes) and one adult male (from Laysail). We assumed tlicse samples were representative of the island at which tlicy were collected because no pups born to known itnmi- grant mothers were included, and very few juveniles are sighted at non-natal locations (Ragen 1995). We also as- sun~ecl these samples were from unrelated animals, although the breeding system in this species is unknown, and if the degree of polygyny is high, son~e of the pups coiilcl be half-siblings. Nevertheless, because monk seal mating occurs in tlic water, the opportunity for inclivicl- ual males to monopolize access to females is limited, and the degree of polygyny is likely to he much lower than in the closely related terrestrial-breeding elephant seals (Boness et al. 1993). 'l'lie probability of sampling related iiiimals was also minimized by sampling individuals from different locations within each island (although the extent of geographic separation varies greatly among sites due to island topography).

Mitochondria1 DNA Sequencing and DNA Fingerprinting

DNA was extracted from tissue samples from 10 individ- rials from each of tlic five islands, following standard protocols (Miniatis ct ill. 1982). Amplifications via the poly- merase chain reaction and generation of single-stranded product for direct sequencing (Sequenase, US Biochcn~i- ~ i l ) were carried out using the mtDNA control region primers described in Kocher ct al. ( 1989; LI 5915) and Mcyer et al. (1990; H 16498). Negative controls for con- tamination were included with each set of samples. A sequence of 559 base pairs was obtained from all 50 seals, including 56 base pairs of the Proline tRNA gene and 303 base pairs of the control region, including all of the "hypervariable" region preceding the "conserved central block" (Slade et al. 1994). An additional 96 base pairs of sequence, for a total length of 455 base pairs, was 01)- tained from some individuals for which both strands were sequencecl and/or tlic sequence was exceptionally clear

Unlike the n~itochondrial sequencing approach, DNA fingerprinting is dependent upon obtaining high quality, non-degraded DNA (see Bruford ct al. 1992). Sample quality was examined by running out a small aliquot of DNA on an agarose gel stained with ethidiutn bromide and visualizing under lN light. Only 22 saniples were found to contain sufficient high n~olecular weight DNA suitable for fingerprinting; tlicsc included 4-7 individuals from four of the five main siibpopiilations. These saniples were digested with l l a v 111 and run on agarose gels at 19 mAinps for 48 hours. Samples from two different islands were run on each gel, with a lane of n~olecu- ; i r weight marker ("Genetic Analysis" ladder, Promega) adjacent to each sample lane. Following transfer of DNA

Conservation Biology Volume 11. No. 2 , April 1997

Kretzmann ct a/. Loii (icnetic Variiibilit}' in Monk Seals 485

fragments from gel to nylon membrane, hybridi~ation was accomplished using c l~en~ i l i~n~ inesce~~ t alkaline phosphatase-labeled 33.6 probe (Cellmark) and "Ge- netic Analysis" marker probe (Proniega), according to the method of Ruth and Fain (1 993).

Fingerprint Analysis

Fragments sized between 2 and 23 kb were scored using an automatic scanner (Scanmaster 3 + , Howtek) and Bio- Image computer software, and band s i ~ e s were deter- mined by reference to the marker lanes adjacent to each sample Line. The similarity index was defined as the number of fragments shared by individuals x and j l , di- vided by the number of fragments detected in both individuals (Lynch 1990). The computer program SIM (version 1.01; Zimmerman 1993) was used to generate within and between island siniilarity indices for individii- ills run on the sanie gel, using a match window of 3 SD (derived from the variation in migration distance for fragments of known size) around each band (see Gal- braith et al. 1991). The SIM program provides variance estimates corrected for the non-indepenclenee of data points in multiple pairwise comparisons, as described by Lynch (1 990), and estimates the extent of population subdivision according to Lynch (1 991):

/',-; = ( 1 - s,,)/(2 - st/, - s,,),

where s,, is the average between-subpopulation similarity (corrected for within-si~bpopiilatiot~ similarity) and s is the average witI~it~-sul>popi~lation siniilarity. This provides a downwardly-biased, conservative estimate of population subdivision (Lynch 1991).

Results

Monk seal control region sequence variation was very' low; of the 359 sites surveyed for all 50 seals, only 2 were variable. No additional variation was detected among the seals for which longer sequences were 01)- tiiined. The 2 variable sites defined three haplotypes (Ta- ble I), one of which (111) was found in a single individual

Table 1. Distribution of monk seal mtDNA control region haplotypes by location.

Island '

H u p l o t y p e I ' t ~ S LAY LLS PHR KlJR Iota1

I 9 7 9 8 1 0 43 I1 1 2 I 2 0 6 Ill 0 1 0 0 0 1

Total 1 0 1 0 1 0 1 0 10 50

l:IÂ¥',S French Private Shoals; LAY, 1,~iysan Island; /./,S, IJsiunski Is- hind; PHR, Pvml and Ilermes Rvvf; K l TI<, Kure Atoll (see I-'nf. I ) .

(the adult male froni Laysan). The sequence for haplotype 1 is available in GenBank (accession number I J59753); haplotypes I1 and 111 were distinguished froni I by transi- tions at positions 58 (in Pro tRNA gene) and 176 (in the control region) respectively. The niost common haplotype (1) was found in 43 of 50 individuals, and the remaining haplotype (11) was found in 6 individuals, representing four of the five principal breeding sites (Table 1). There- fore, the control region sequence data provide n o evidence for genetic differentiation anlorig monk seal subpopulations.

I'he combination of the restriction enzyme Ilav 111 and alkaline pl~osphatase-labeled 33.6 probe yielded coni- plex nionk seal DNA fingerprints, with 15-3 1 scorable bands per individual (Fig. 2). Comparisons were limited to two pairs of adjacent islands run on the sanie gel: Lay- ran versus Lisianski and Pearl and Hernies versus Ki~re. Mean within-island similarity estimates ranged from 0.49 to 0.73, and (with one exception) the similarity values for individuals from different islands were lower than the within-island values (Table 2). Although the distributions of band-sharing values within and between islands overlapped, and similarity values between sites were quite high, the differences between these distributions were large enough to yield signific;mt ( p < 0.01) estimates of population subdivision in both comparisons, with estimated/", values of 0.20 and 0.13 (Table 2).

Discussion

Evidence for Low Genetic Variability Within Subpopulations

D;ita from both miDNA sequencing and niultiloci~s DNA fingerprinting support the hypothesis that monk seal genetic variability is extremely low. Even the cheetah, shown to be depauperate in genetic variation at a variety of nuclear coding loci, exhibited "moderate levels of genetic diversity" in both DNA fingerprint profiles and mi- tochonclrial DNA restriction fragment analysis of the whole genome (Menotti-Raynioncl & O'Brien 1993). In contrast, even within the niost variable region of the mi- tochonclritil genome, we found only three control region l~aplotypcs among 50 individuals representing the entire range of the species, and a very low percentage (0.6) of variable sites in this "hypervariable" region (Table 3).

Monk seal control region sequence variation can be cli- rcetly compared to homologous sequence data obtained froni other pinniped species that have been similarly surveyed (Table 3). This analysis shows that the northern elephant seal is the only other species with a similar paucity of genetic variation in this region (both in terms of haplotypic and nueleotide diversity). The northern elephant seal is a close relative of the nionk seal, and has a similar history' o f human exploitation leading to a well- documented population bottleneck. It has been shown

( , o ~ i ' i u v . i l i o n Biology Volume I I . No 2 . April \W7

486 Low Genetic Variability in Monk Seals Kretzmann et al

Figure 2. Example of a DNA fingerprint for Hawaiian monk seals from Pearl and Hermes Reef. The darker lanes contain molecular weight marker. Fragment sizes are shown at left. DNA was cut with Hae 111 and hybridized with alkaline phosphatase-labeled Jeffreys probe 33.6.

to be depauperate in genetic variation using several approaches (allozyme electrophoresis, Bonnell & Selander 1974; DNA fingerprinting, Lehman et al. 1993; and mtDNA sequencing, Hoelzel et al. 1993).

The fingerprint data must be considered preliminary due to the small number of samples that proved to contain suitable high quality DNA. Nevertheless, for carni- vore populations similarly assessed with Jeffrey's multilocus fingerprint probes, band-sharing values above 70% for "unrelated" individuals have been reported only within populations that are small and isolated and/or highly inbred (e.g., Channel Island fox, Gilbert et al. 1990; Isle Royale wolf, Wayne et al. 1991; northern ele-

Table 2. Summary of DNA fingerprint results* for Hawaiian monk seals from four islands.

Island

Result Pearl and component* Luysan Lisianski Hermes Kure

N 7 4 6 5 Mean n 18 16 27 24 Mean s.,, 0.65 0.49 0.71 0.73 Mean si, 0.51 0.64 f 'st 0.13 0.20 C.L. 0.07 0.02

N , number of individuals; n, number of bands scored; sw and sb, similarity indices for individuals within and between islands, respectively. The ps , and C.L. (99% confidence limits on f'.) were cal- culated based on Lynch U W ) .

phant seal, Lehman et al. >1993). In contrast, outbred mammalian populations typically show similarity values between 20% and 60% (e.g., Ruth & Fain 1993). DNA fm- gerprint similarity values are well correlated with known inbreeding coefficients in domestic poultry (Kuhnlein et al. 1990), and inbred populations of both laboratory mice (Jeffreys et al. 1987) and humans (Bellamy et al. 1991) show high band-sharing values relative to outbred controls. The similarity estimates obtained in this study are therefore consistent with local inbreeding within at least some monk seal subpopulations.

Problems with DNA Fingerprint Interpretation

Interpretation of our DNA fingerprint results in the context of other studies is somewhat problematic. Band- sharing estimates for the same individuals may vary sig- nificantly depending on the particular enzymelprobe combination used (e.g., Georges et al. 1988; Hanotte et al. 1992) and, because there is no standard method for scoring fingerprints, varying match windows may sub- stantially alter band-sharing values. In addition, samples of poor quality are likely to yield misleading results. De- graded DNA will be missing the high molecular weight fragments which tend to show the highest variability (Fig. 2), thereby biasing similarity values upwards. By limiting our analyses to the subset of samples with little or no DNA degradation, we avoided this pitfall.

Inbreeding as a Threat to the Population

In contrast to the monk seal, the northern elephant seal demonstrates ongoing rapid population growth and is now estimated to number over 150,000 individuals (see Stewart et al. 1994), from a post-bottleneck population that probably numbered fewer than 100 (see Hoelzel et al. 1993). These seal species share low genetic variability but have responded very differently following a severe reduction in numbers. Therefore, it might be ar-

Conservation Biology Volume 11, No. 2 , April 1997

Table 3. Intraspecific nitDNA control region sequence variation for pinnipeds.

No. of'h~~/)loIy/)es/?~o. ofindividua1.s No. of /~c/rit,~/)lv .sites/ Va~,ic//?le sites Species (no. of"/)o/)~ ~Uitioiis ") totc~l no. sequc~iicetl (%)

Hawaiian monk seal Northern elephant seal" Southern elephant seal" Harbor seiil" California sea lion' Steller sea lion"

gued that the two differ markedly in their sensitivity to inbreeding, or, alternatively, that genetic variation per se is not a good predictor of the ability of an endangered species to recover. Inbreeding costs are known to vxy widely among mammals (Ralls et al. 1979; 1988), but the extent to which the monk seal may suffer from the delete- rious effects of inbreeding is unclear. Inbreeding depression is often manifested as reduced fertility and/or poor juvenile survival in other mammals (Ralls et al. 1979; Brewer et al. 1990; Laikre & Ryman 1991). While female monk seal fecundity is quite low compared to that of other seals (Johanos et al. 1994), juvenile survival in this species can be quite high: 80-90% in the first year and 85- 98% for young seals over age one (dilmartin et al. 1993).

High DNA fingerprint band-sharing values between mated pairs of Puerto Kican Parrots (Brock & White 1992) and Great Reed Warblers (Bensch et al. 1994) were associated with reproductive failure. A similar effect is therefore a plausible explanation for the low reproductive rate of the Hawaiian monk seal. Patenaude et al. (1994) recently concluded that inbreeding depression could be a factor in the failure of an endangered be lu~a whale population to recover, based on high band-sharing values relative to a healthy reference population of the same species. Inbreeding depression has also been invoked as an explanation for poor reproductive performance in captive cheetah populations (O'Brien et al. 1985), but other studies indicate that non-genetic factors account for most cheetah mortality in the wild (Caro & Laurenson 1994; Laurenson et al. 1995). Similarly, ecological factors such as lack of suitable habitat and a decline in food resources (Polovina et al. 1994) may pose a more important threat to the monk seal population than do genetic factors.

Although a lack of genetic diversity may not always be of foremost concern in conservation efforts, the potential importance of low genetic variability for increasing the vulnerability of a species to infectious disease is often cited (e.g., Bonnell & Selander 1974; O'Brien & Kver- mann 1988). This effect may be especially important in small populations. Marine mammal epizootics are not

uncommon; a phocine distemper virus killed nearly 18,000 common seals in Europe in 1988 (Harwood &

Hall 1990). A similar outbreak among Hawaiian monk seals could potentially drive the species to extinction. Furthermore, recent work by Kellcr et al. (1994) demonstrated a clear survival bias against inbred individuals following a song sparrow population crash, and Jimenez et al. (1 994) reported that inbreeding had a more detrimen- tal effect on mice introduced into a natural habitat than on those maintained in the laboratory. Thus it appears that previously undetected inbreeding depression may be manifested in the face of an environmental challenge. Even an apparently highly successful species like the northern elephant seal might therefore become vulnera- ble, due to depressed levels of genetic variation, should environmental conditions change markedly.

Evidence for Population Subdivision

'She two genetic marker systems used in this study provide conflicting evidence regarding population subdivision. Our extensive mitochondrial control region sequence data set does not indicate genetic differentiation among si~bpopukitions. Indeed, the presence of a relatively rare haplotype on four of five islands (see Table 1) supports the idea that the monk seal population is panmictic. In contrast, the preliminary results of the fingerprinting study indicate that individuals from the same island are genetically more similar than are individuals from adjacent islands. This suggests that some degree of monk seal subpopulation differentiation exists, and that gene flow among the islands has been limited.

Population genetic structure (and local inbreeding) in other species lias been inferred from non-overlapping within- and between-subpopulation similarity values (e.g., IIoelxel & Dover 1991; 'l'riggs et al. 1992). Al- though the ranges of similarity values between monk seal suhpopulations overlapped those within siibpopula- tions, the distributions were different enough to allow detection of significant genetic stn~cture in the population as a whole. We concluded that two adjacent pairs of

islands (Kure vs. Pearl and IIerines Reef, and 1.isianski G. Ainato, S. Safiudo-Wilhelniy, and an anonymous re- vs. Laysan; Fig. 1) were genetically distinguishable from viewer. This work was partially supported by a Ifniver- one another, with Itst values of 0.20 and 0.13, respec- sity of California Seed Fund Grant to DPC and MBK, and lively (Table 2). These are probably minimum estimates by NSF grants (DEB-8918027, CSR-9107838 and BSR- of the proportion of genetic diversity distributed among 9 1 19867) to AM. Computer analysis of DNA fingerprints s~~bpopulations, because more distant islands in the chain was performed at the US Fish and Wildlife Forensics Lab- are likely to differ even more. oratory in Ashland, OR.

Implications for Conservation

The genetic data presented here liave several possible implications for the management practice of translocating seals among islands. Movements of individuals among subpopulations is often opposed due to the risk of disease transmission (e.g., Wilson et al. 1994; Brower et al. 1995). This may be an issue of particular importance for the monk seal, in light of the paucity of genetic variation observed in this species. Also, if monk seal subpopulations are indeed genetically distinct, then increasing the effective migration rate among islands may have the potential to reduce genetic variation retained in tlie population as a whole. However, the genetic differentiation revealed by DNA fingerprinting was not supported by the mtDNA sequence data and may not reflect biolog- ically meaningful differentiation among islands, because differences in coding DNA will recover much more slowly following a population bottleneck.

J contrast The results presented here for tlie monk se* 1 markedly with tlie DNA fingerprinting results reported for an endangered Hawaiian bird, in which two subpop- illations were each genetically quite diverse but not at all differentiated from one another; in this case pro- posed translocations from the larger to the smaller subpopulation were clearly favored by the genetic data (Fleischer e l al. 1994). For the monk seal the survival benefits associated with moving animals to a location

Literature Cited

Allernlorf, 1'. W 198.3. Isol;ition, gene How iiml gcnelic tliffcrcntialion iiiiong populations Isages 51-05 in (:. M. Selionewalcl-Cox, S. M. Chambers, IS. MacBrydc anil W. I.. ' l ' liom;~~, cdilors. (ienctics and ~onserv'.ition: a refcrctiee for managing wild minr.il iind plant pop- illations. Itenjamin/C~imiiiings, Menlo Park, California.

Itearhiore, J. A. 1983. Exlinetion, survival, and gem-lie variation. 'ages 125-151 in (:. M Sclioiiew;iId-Cox, S . M. (.li:inilx~s, It. M;~clirydc am1 W I.. Thomas, editors. Genelics and conservation: a reference for m;in:iging wild animal and plant [iopiil:ilions. Ikn- jamin/Cumniings, Menlo Park, California.

liellamy, It. J. , (:. 1: Ingleliearn, 1. K . Jalili, A. J. Jefkys , ancl S S. Bhatta- cliaryi. 199 1. Incrcasecl b;incl s1i;iring in DNA fiiiger1irints of ;in in- hnxl human population. Human denctics: 541 -W.

Itenscli, S., I). Has~-lc~uisl, ;ind I' von Sch:int/. 1991. (iewtic similarity Ix'tween parenis predicts h;itcliing failure: iionincesti~oi~s inlireed- ing in llie great r m l w:irblcr? Evoliition 48:317-326.

Kickham, J. W., ). (;. Patton, and I' It 1.ougIilin. 1990. High variability for control-region scc[i~cnces in :I m;irinc m;~mm;il' implicalions for conserv:ition and biogeography of Slellcr s u lions (Enttzetopin^./'~/- hdliis). Joiirmil of M'.imm.ilog~ 77:95-108.

Hirky, C. W., P. Fiwst, ;ind '1'. Mi~rny;ini:i. 1989. 0rg;inelle g ~ 1 e cliversity under migration, mutation and ilrift' equilibrium expcct;~lions, approach to equilibrium, (.Tfecls of lieteropl:ismic cells, and coni- p;irisoii to nuclear genes (icnetics 121:61.1-027.

lioness, I ) 1.. W . 1). liowcn, :ind J. M. I'rancis. 1993. 1mplic:itions of f)NA fingerprinting for iiiiiting systems ;inif reproductive str;Uegies of pinnipeds. Symposium of tlie Zoologieiil Society of London 66: 01 -93.

lionnell, M . I.., and It. K . Sekinder. 1971 Hlepliant seals: fern-tic variation :ind ne;ir extinelion. Science 184:908-909.

where food resources are more plentiful and tlie pOssi- 13rewer, 13. A , R (: l.:~cy, M. 1. Foster, anc1 G. Aklks. 1990 lnliweiling depression in insi1I:ir :unI cenlriil popi~I;~lions of l'eriiinyscus niice.

ble advantages of relieving local inbreeding argue in fa- Journ:~l of Hemlity 81:257-260.

vor of the translocation program. Although inbreeding l h c k , M , K , .lnil 1%. N . white. 1992. Application of DNA fingerpriiiting - - . .

depression has not been conclusively demonstrated in to the recovery progrilm otthe encli~ngcreil I'uerto Rican parrot. ~ r o -

tins species, the evidence for low genetic variabilitv and ceedings ot the ~ . i t i~n . i l Ai-.ldeni) ot SCIC ines 89: l 1 121- 1 1 125

local inbreeding is strong. ~h~ monk seal,s low fecu,,. lirower, 1.. I' , ct ;%I. 1995. On the il;ingcrs of interpopiil:ition:il tnuisfei-, of mon:ircli bu~tcrflies. IiioScience 4 5 5 10-5 14

dity represents a real threat to its recovery, and a genetic liruforil, M, W , 0. H;inottc, J I-'. Y. Brookliekl, ;ind '1'. Iturkc 1992. Sin-

basis for this poor reproductive performance is certainly gle-10~11s am1 miiItiIocus DNA tingcrprintiiig. P ; I ~ O 225-269 in A. R. possible.

Acknowledgments

Hoelzel, editor. Moleciiliir genelie ;~n;~lysis of popiilatioiis: a praclic:~l :ippro;~eh. Oxford University Press, Oxford, Uniteil Kingdoni

Caro, '1'. M. , and M . ti 1.a~irensoii 1994. Ecological am1 genetic factors in conservation: a cautionary tale. Science 2631%-480.

Croniii, M. A. 1993 Mitoclionclri:il DNA in wildlife taxonomy and con- st~v;ilion biology: ei~iition;i~y notes. Wildlife Society Ik~llelin 21:

We are grateful to B. Rice and M. Soule, who provided 539-348. laboratory facilities and helpful discussion throughout ~:k-isclier, R. (:., (:. I.. 'l'arr, anill'. K . ~ n i t t 1991. (icnctic siructiire ;ind

the course of this study, to P. Ritchie, who provided as- mating system in the palila, ;in enilangcreci Hawaiian honey-

sistance with the 1'C.R and DNA sequencing protocols, crcelm, :is iisscssed hy DNA fingerprinting. Molecnl:ir Ecology 3: 383-392.

:lllci to ntlmmxls colle:tg~les at the Nati(m1 Marille Fis11- l~r~l l l l~ l l~ l i l l , 1995, 1111 ,reei lillg 2111cl cxtiilctiOil: tllresllOl~l effect.

cries Service who conducted tlie field sampling. The tion on KioIogy 9-.792-799. nlanuscript was improved by suggestions from A. Dizon, d.~lbr.~ith, 11. A,, 1' 1'. ~io;ig, 1~1. I.. ditibs, and 1% N. White. 1991 Sizing

Conservation biology Volume I I, N o 2, April 1997

bancls o n :iiitor:~cliognims. :I study of precision for scoring DNA fingerprints. HIectrophoresis 12:210-220.

Gorges, M., A,-S. I.equarrc, M. Castelli, R. Hanset, :ind G. Vassart. 1988. DNA fingerprinting in domeslic anim:ils using four differen1 minisiitcllite probes. Cytogenetics and (;ell Genelics 47: 127- 13 1.

(- ~ciroclette, .. T., iincl W. C;. Giliiiiirtin. 1090. Ik-niograpliic coiise-

qiiences of changed pupping and hauling sites of tile Hawaiian monk seal. Conservation Biology 4:42.3-130.

Gilbert, I). A., N. I.eliniiin, S. J O'ltrien, iinil R. K. Wayne. 1990. Ge- in-tic fingerprinting rellecfi population clifferenti:ition in the C.ili- forniii Channel Island fox. Nature 344:764-767,

Gilniartin, W. G., '1'. C. Joli:inos, and I.. 1.. Hberharclt. 1 005. Survival for the Hawaiian monk seal (Mottnrlnis schm~ittslci?tdi). Ma-

rine Mammal Science 9407-120. (iilmartin, W. G. , am! I.. I.. Ehcrliardt. 1995. Status o f the Hawaiian

monk s ~ i l (~I'foinichns schauiittslu?ttli) population. Qint~liiin Jour- n:il of Zoology 73: 1 185- 1190.

Gilpin, M . H., and M. E. Sonic'. 1986. Mininium viable popukilions: pro- cesses of slxcies extinclion. Pages 19-34 in M.E. Soulc', editor. Conservation biology: the science of sc:ircity and diversity. Sintitier Associtites. Sunclerltind, Mtiss:iclinsells.

ILinotte, 0.. M , W. ltruford, and 1' Iliirke. 1992. Multilociis UNA fingerprints in g:illiniicco~is birds: goieral approach and problems. Hercdily 66:191-195.

Harwood, J., and A. Hall. 1900. Mass niortality in marine mammals its implie:itions for popukition dyntiiiiics iind genetics. Trends in Ecol- ogy and Evolution 8:254-257

Hoelzel, A. I<., iind G . A. Dover. 1991. Gn~et ic differentiation between sympatric killer wliiilc populations. Heredily 66: 191 - 195.

Hoelzcl, A. I<.,l. lLillcy, S . ) . O'ltrien, (:. (hiipiign;i, I' Arnbom, 13.1. 1.e Itociif, K. Rails, :inel G. A. Dover. 1993. Elepliam seal genetic v:irki- tion :ind the use of simulation models to investigate historical pop- illation bottlenecks. Journal of lh-rcdity 84:113-449

~ f f r e y s , A. J., V. Wilson, and S. I. Thein. 1985. l~Iyl~'~'viirititile 'minis:it- cllite' regions in Iiiiniiin DNA. Nature 314:67-73.

Jeffreys, A. J., V. Wilson. I< . Kelly, 13. A. Tiiylor, : i d G. Itulfield 1987. Mouse DNA 'fingerprints:' an:ilysis ofcliromoson~e IoealixiUion and geriii-line stability of Iiypc~viiriabIe loci in recomhiniint inbred stniins. Nucleic Acids Rcs~irch 15:282.->-2836.

Jinic'ncz, J. A , , ti. A. l~IugIies, G. Akiks, I.. Gniliiim, and I<. C;. 1;icy. 1001. An experimental study of inbreeiling depression in a natural lialii- tat. Science 266:271-273

Joli:inos, '1'. (;., I t 1.. Ik'ckcr, and '1'. J . Ragen. 1994. Anniiiil reproiltic- tive cycle of the fem:ile Hawaiian monk s ~ i l (Mottuchus schi~iitii- slintcli). M:irine M~nim:~l Science 10: 13-30.

Johnson, A. M , R. I.. DeLong, C. H. Fisciis, :inil K. tienyon. 1 082 l'opii- lation status of the Hawaiian monk seal (Moiuichus schuuinsliiiicl~> 1978. Joiirn;il of M2nim;ilogy 6 5 4 15-421

Johnson, A. M., :ind E. Kricller 1983. Interiskind movenienl of H:iw:ii- ian monk seals. HIep:iio 44:43-45.

Keller, I.. I-'., P. Arccse, J. N. M. Smith, W. M . Hoc1i;~-likii, :inel S. (;. Steariis. 1991. Selection against inbred song sparrows during a nat- unil popiil:ition bottleneck. Nature 372:356-557.

Kenyon, K. W. 1972. Man vers~is the monk sc.11. J o ~ i r i ~ i l of Munimiilogy 53:687-696.

tienyon, K. W., and 11. W. Rice. 1950. Life liistoty of llic Hawaiian monk se:il. Piicific Science 1 3 2 15-252.

Kocher, '1'. 0., W. K. Thomas, A. Meyer, S. V . Eilw:irds, S. l'iiiibo, F. X. ViIl:iblaneii, and A. (:. Wilson. 1089. 11yn:in'iics of niitocliondrial DNA evolution in ailim:ils: iimplific:ition and sec1iiencing with conserved primers. Proceedings of llie Nii1ion:il Aciiclemy of Seiences 86:61%-6200.

Kuhnlein, U., 1). X;i(.lworny, Y. Ikiwc, I<. W. k~irfiill, :ind). S. Gavor;~. 0 9 0 . Assessment of inbreeding by DNA fingerprinting: clcvelop- merit of ;I calibration curve using clcfinci-1 strains of chickens. dc- nctics 125:161-105.

Lacy, R. C. 1087. Loss of genelic cliversity from nitin:igeil popukitions:

intenieling effects of drift, imn~ig~ition, selection, iincl poptil:ition subdivision. Conscrv:ition Itiology 1: 143- 158.

L:iikre, I. , ancl N. I<yni;in. 1991 Inbreeding clepression in a captive wolf((,>i}iis ln/)iis) pop~iliition. Consa'v.ition lliology 5~35-40.

Lunik-, R. 1988. (ienetics am1 demography in biological conservation. Science 24l:l455- 1460.

I.:iiirei~soi~, M . K., '1'. M. C.iro, 1'. Gros, iincl N. Wielcbnowski. 1995. Controversiiil c1ieet;ili.s~ Nulure 377:.592.

I.eIim:i~~, N., I<. K. Wiiync, iind It S. Slewart. 1993. Comp:irativc levels of genetic variability in harbour seals am1 northern elcpliant seals ;is cletermineil by genetic fingerprinting. Symposium of lhe Xoolog- ic:il Society of London 6 1 9 - 0 0 .

Lynch, M . 1990. The siniikirity index iind DNA fingerprinting. Molecii- :ir Itiology and Evoliition 7:478-181.

I.yiicli, M. 1991. Analysis of population genetic structure by DNA Fin- gerprinting. Pages 113-120 in 1'. Itiirke, (;. Dolf, A. J Jcftrcys, iincl R . Wolff, etlilors. ONA fingerprinting: approiichcs ;ind tipplic;ition.s. I l i rk l~i~~ser Verl:ig, It.iscl, Switxerliinil.

Malilonado, J . E., F. Ona lkivila, It. S. Stewart, E. (;offen, and I<. ti. W:iyne. 1995. Intraspccific genetic iliffercntiiition in (:;ilil'orniii sea lions (Zn1ol)hu.s i'~~lifoniiiiiiii.s) from soiill~ern diliforniii and the Gulf of California. Marine Mammal Science 1 1:46-58

Maniatis, 1' , E. 1'. h'ritsch, and J. Sambn)ok. 1982. Molecular cloning: :I

l:iboratory mtinn:il. Pisst cilition. Gokl Spring Harbor Lalionitory 'ress, Cokl Spring Harbor, New York.

Menotli-I<:iyinomI, M . , and S. J. O'Hricn. 1993. Dating tlie genetic hot- ticneck of tlie African clicct:ili. Procecilings of the National AeiicI- eniy of Sciences 90:3 172-3 170

Merola, M. 1991. A re;issc-isiiient of homozygosily and the c:isc for in- weecling depression in the cheetah, AchKinyx jiibaliis: impliea- lions for conservation. C.oncrv:ition lliology S:%1-971.

Mcyer, A , , '1'. 0. Kocher, 1'. H;is:isibw;iki, iind A (:. Wilson. 1990. Mono- pliyletie origin of Luke Victoria ciclilid fishes siiggesleil hy niitocliondrial DNA scc1ucncci. Nature 347:550-555.

Mcycr, A . 1994. Sliortcomings 01' tile eytocliroiiic b gcnc :is a molccu- lar m:irker. Trcruls in Ecology :ind Lvolution 9:278-280.

Morit.', C. 1994. Applic:itions of niiloehorulritil DNA analysis in eoiiser- valion: :I critical review. Moleciiliir Ecology 3:401- 11 1

O'Hrien, S. ). 1994. The cliect;ih's consemition controversy. Consi~v.i- lion Itiology 8:1 153-1 155.

O'ltrien, S . J., ancl J . 1'. Evcrni:inn 1988. Inte~ictive influence of infec- tioiis clisease :incl geiietic diversity in n;ilur:il popiil:ilions. Trends in Ecology ancl Hvol~ilion 3:251-259.

O'ltrien, S. J., M E. Roelke, I.. M;irker, A Newman, C:. A. Winklcr, I). Mcltxcr, 1.. Colly, J 1: Evcrmann, M . Il~isli, and I) . E. Wililt. 1985. Genetic basis for species viilneniliilily in the cheetah Science 227: 1428-1 151.

l':itena~iilc, N. J . , J . S Qiiinn, 1'. Hekind, M. Kingsley, am1 It. N White. 1991. (ienetic variation of the St. Lawrence belug:i whale popul:ilion assessed by DNA fingwprinting Molecular Hcology 3:575-581.

I'olovina, J. J , (i. '1'. Mitclium, N. E (irahaiii, M. P. Craig, E. E DcMar- tini, iunl E. N. Flint. 1994 Pliysiciil and biological conseqiienccs of :I climate event in the central North Pacific. Fisheries Oceanogra- pliy 315-21.

Ragen, '1'. J . 1903. Stalus of the Hawaiian monk seal in 1992. Report H-93-05 Soiithwfit Fisheries Science Center Honolulu Laboratory, Honolulu.

Ralls, K., K. Uriigger, and J Killo~i 1979. Inbreeding and juvenile iiior- tality in smiill poptihilions of iingulatcs. Science 206: 1101- 1 103.

Ralls, K., J . 1). Ikillou, and A. Tenipleton. 1988. Eslimates of Ietliiil ecl~iivalenis am1 the cost of inbreeding clepression. (:onserv;ilion Iti- ology 2:185- 193

Ruth, ). I.., and S. R. Pain. 1993 Tile ' i i i ~ l i v i i l ~ i t i l i ~ i t i o n ' of large North Aiiicriciin mammals. Images 129-450 in S. I). J. Penii, R. (:Iiak- nihorty, ). '1'. Epplen, and A. J. Jeffreys, editors. DNA fingerprinling: ,sl;Ã‡ of the science. Kirkliiiu.scr Vcrhig, lkisel, Switxerl;iiiiI.

Sl:ide, R. W., (:. Moritz, ;inel A. Heiilem:in. 1991. Multiple nucl~ir-gene

(:oi~~c~viition J-iiolog Volume I I , No 2 , April I W

phylogenies: :ippliciition to pinnnipeds iinti comptirison with a mi- tocliondrial gene phylogeny. Molecular Itiology ii11cl Kvolution l l: 341-356

Stanley, H. F., S. Case),. J . M. C.iriialian, S. Goodman, J. Harwood, and R. ti. Wayne. 1096. Worlclwiclc piitterns of mitoehonclrial DNA tlif- fcren1i:ition in the harbor seal (l'bocii ritiiliiiu). Molecular Itiolog), :ind Evolution 13:368-382.

Stewart, It. S.. 1'. ti. Yochcm, H . 11. Hulx-r, R. 1, I)eI.ong, lt J . Jameson, W. J. Sycle~ii:in, S. (7. Allen, and It. J . 1.e Koeiif. 1991. l'iiges 29- 1% in It. 1. 1.c Hoeiif am! It. M. Laws. editors. Hlcpliiinl seals' population ecology, beIi:ivior and physiology. University of Ciiliforniii Press, Itcrkclcy.

Triggs, S J , M . 1. Williams, S. J Marsliall, :inel (;. K. Chanilx-rs. 1992 Genetic striieliire of blue {luck (llyiin'iui/ai~;ii;.s niuliito~piichos) populations reve:ilcd by DNA fingerprinting. The Auk 109:HO-89

Varvio, S.-I.., It. Cliakrahorly, :liitl M . Nci. 1986. (;enctic v:iri:itioii in subdivided popul:ilions ;ind conscrv.itio~i geiietics. Hewdity 57: 189-198.

Vrijenhoek, R. (:. 1994. (icnetic tliversily : i~it l fitness in \mall pop~ilii- lions. l'iiges 57-53 in V. Loesclicke, 1. 'l'omi~ik, ;ind S. K. Jiiin, wli- tor\. Go~iser~iition genetics. ltirkliikiser Verltif Ikisel, Switzerland

w~y~ie . It. K . , et ;il. 1991. Coascrvatioii genetics ol'llie entlangered Isle ltoyale grey wolf. (;onser~.ition Itiology 5: 11 -5 1 .

Wilson, M , H., (;. I t Kepler, N I-'. It. Snyclcr, S. R. Derrickson, 1'.J Dein, J W. Wilcy, J M. Winitierle. Jr., A E. l.ug0, I ) I.. (;r;ili:im, and W. 1) Toone. 199 t Piierlo Iticaii 1xirrot.s initl potential liinitiitions of the mct.ipop~~l.i~ioii appro.n-'h to species co~~serv;itioti. ('.onser- v:ition ltiology 8: 1 I I- 123

/.i~iinierma~i, R (;. 1993. Progniiii SIM, version 1 .O1 llnivcrsity of Ciili- lorniii Regents, 1.05 Angelcs

Conscrv:~tion Biology

Volume 1 1 , No 2, April 1997

Documents

Low Genetic Variability in the Hawaiian Monk Seal