634

Wildlife Society Bulletin 2002, 30(2):634–643 Peer edited

Genetic consequences of hunting: what do we knowand what should we do?

Richard B. Harris, William A. Wall, and Fred W. Allendorf

Abstract Possible evolutionary consequences of sport hunting have received relatively little con-sideration by wildlife managers. We reviewed the literature on genetic implications ofsport hunting of terrestrial vertebrates and recommend research directions to address cur-rent uncertainties. Four potential effects can be ascribed to sport hunting: 1) it may alterthe rate of gene flow among neighboring demes, 2) it may alter the rate of genetic driftthrough its effect on genetically effective population size, 3) it may decrease fitness bydeliberately culling individuals with traits deemed undesirable by hunters or managers,and 4) it may inadvertently decrease fitness by selectively removing individuals with traitsdesired by hunters. Which, if any, of these effects are serious concerns depends on thenature and intensity of harvest as well as the demographic characteristics and breedingsystem of the species at issue. Undesirable genetic consequences from hunting havebeen documented in only a few cases, and we see no urgency. However, studies specif-ically investigating these issues have been rare, and such consequences require carefulanalysis and long time periods to detect. Existing information is sufficient to suggest thathunting regimes producing sex- and age-specific mortality patterns similar to those occur-ring naturally, or which maintain demographic structures conducive to natural breedingpatterns, will have fewer long-term evolutionary consequences than those producinghighly uncharacteristic mortality patterns.

Key words alleles, effective population size, evolution, gene flow, genetics, heterozygosity, hunting,selection

Commentary

Wildlife managers have historically placed greatemphasis on demographic issues and relatively lit-tle on how hunting influences genetic characteris-tics or the evolution of populations (Rhodes andSmith 1992). However, speaking as hunters as wellas biologists, we value hunting as an experience ofthe wild, which in our view is linked inextricably tothe forces of natural selection that have producedour native species. Few hunters would find interestor fulfillment in stalking animals housed in a zoo;

similarly, we believe our descendents deserveopportunities to interact with species shaped prin-cipally by their native environments, rather thanartificially molded in adaptation to human desires.In our opinion, integrity of a wild species’ genepool deserves respect similar to that accorded tomaintenance of its natural habitat.

Unlike in fisheries, however (e.g., Sutherland1990, Stokes et al. 1993, Law 2000), North Americangame managers have paid little attention to the

Commentary • Harris et al. 635

long-term effects hunting may have on the geneticmakeup of game species. This is surprising becausesport hunting alters population density, sex ratio, andage distribution (Wall 1989, Ginsberg and Milner-Gulland 1994, Solberg et al. 2000), all of whichpotentially influence the genetics of populations.

We addressed four questions relating to commonsport hunting practices and their long-term geneticconsequences: 1) Can hunting alter natural patternsof gene flow among demes? 2) Can hunting lowergenetic variation, through increasing genetic driftcaused by reduction of effective population size?3) Can deliberate selection against traits viewed asundesirable by hunters or managers reduce fitness?and 4) Can unintentional selection pressures, usual-ly arising from management based on demographiccriteria alone, have unintended consequences? Wereviewed the existing literature to examine thesequestions, and provide some interpretations andsuggest areas where further research is needed.

Hunting and gene flow Genetic procedures recently described genetic

differentiation within populations that appear mor-phologically uniform. Localized genetic differentia-tion has been documented for many huntedspecies, including pronghorn antelope (Antilo-capra americana, Lee et al. 1989), mouflon (Ovisgmelini, Petit et al. 1997), red deer (Cervus ela-phus, Strandgaard and Simonsen 1993), red fox(Vulpes vulpes, Frati et al. 1998), ring-necked pheas-ants (Phasianus colchicus, Warner et al. 1988,Robertson et al. 1993), and bobwhite quail (Colinusvirginianus, Nedbal et al. 1997). In white-taileddeer (Odocoileus virginianus), considerable differ-entiation exists among neighboring subpopulationson a genetic level (Sheffield et al. 1985, Scribner et al.1997). It appears that local differences in white-tailed deer are maintained by philopatry amongfemales, but subpopulations are prevented frombecoming too subdivided by male dispersal(Mathews and Porter 1993, Purdue et al. 2000; seeCronin et al. 1991 for similar findings in mule deer,O. hemionus). If males become rare through hunt-ing, one concern is that the existing level of geneflow may be reduced further, strengthening popula-tion differentiation at the expense of genetic vari-ability within localized demes (Ellsworth et al. 1994).

Conversely, for species that are naturallyphilopatric or territorial, local genetic adaptationsmight be lost if gene flow among demes isincreased due to the alteration of social structure

caused by hunting. Kurt et al. (1993) examinedpopulations of European roe deer (Capreoluscapreolus) living in both forested and open habi-tats. In forest habitats, male roe deer were territo-rial, controlled access to a number of females, andvariance of reproductive success among males washigh. In open habitats, males were more migratory,a larger proportion of adult males succeeded inbreeding, and genetic mixing within populationswas greater. These differences evidently reflectedadaptive responses to the variability of environ-mental resources, and were maintained locally bylow levels of gene flow. However, under high hunt-ing rates, gene flow between the 2 social systemsincreased, and this differentiation began to breakdown.

Similarly, among territorial greywing francolin(Francolinus africanus), Little et al. (1993) foundno difference in heterozygosity (H, the percentageof loci that are heterozygous in an average individ-ual) between hunted and unhunted populations,but higher levels of inbreeding in unhunted popu-lations. They concluded that any reduction in Hcaused by lower population size was compensatedby greater gene flow within the hunted population.The net effect on H was neutral (i.e., higher migra-tion rates were balanced by fewer potentialmigrants), but hunting clearly had contributed to abreakdown in the usual territorial structure.

Frati et al. (2000) interpreted lower genetic vari-ability among unhunted populations than huntedpopulations of red fox in Europe as reflectingchanges in fox social structure following the loss oflarger predators. Historically, with the presence ofwolves (Canis lupus), leopards (Panthera pardus),and lynx (Lynx lynx, L. pardinus), fox social struc-ture was flexible and outbreeding common. Theysuggested that hunting, by increasing turnover anddecreasing inbreeding, could partially mimic theeffects of predation pressure under which foxeshad evolved.

Hogg (2000) found that mid-ranking malebighorn sheep (Ovis canadensis) from an unhunt-ed population made temporary migrations duringrut to an adjacent hunted population. These ramsfaced less competition for mates from the relativelyfew high-ranking males in the hunted populationand enhanced their breeding opportunities by mov-ing. Here, gene flow from one deme to another wasagain increased, but in this case the probable deter-minant was not hunting per se but rather theabrupt contrast in density of older, larger-hornedmales between adjacent demes.

636 Wildlife Society Bulletin 2002, 30(2):634–643

The available evidence suggested to us that alter-ations in naturally occurring patterns of gene flowwould seem possible from any type of hunt; somelevel of social disruption must accompany anyremoval of individuals. We found this troublesomeonly when locally adapted gene complexes werecompromised by hunting-induced gene flow,where gene flow would otherwise be discouragedby social behaviors. As the red fox example sug-gested, a hunting-induced increase of gene flowamong adjacent demes may help mitigate otherman-made reductions of gene flow.

Hunting and genetic driftGenetic drift is the random change in gene fre-

quencies caused by sampling (via sexual reproduc-tion) from a finite population. Genetic drift occursin all populations, but its effects become pro-nounced only if effective population size (Ne) issmall. Ne is the number of individuals in an idealpopulation expected to lose genetic variation at thesame rate as the census population (N) (Wright1969, Harris and Allendorf 1989). With small Ne, His expected to decline, rare alleles are expected tobe lost, and alleles may become fixed regardless oftheir effect on fitness.

Relevant questions were whether Ne of huntedpopulations might be small enough for drift to be alegitimate concern, and whether hunting regimesfurther reduce it. Concern about small Ne wasexpressed for introduced herds of bighorn sheep(Fitzsimmons et al. 1997) and upland birds (Little etal. 1993). Ryman et al. (1981) simulated moose(Alces alces) and white-tailed deer populations, cal-culating Ne based on various approximations fromdemographic statistics. They found that Ne waslikely to be much lower than N, even in a popula-tion exposed to no selective hunting. Ne/N ratioswere additionally reduced under most hunting sce-narios simulated. Although reductions in Ne/Nunder hunting were not large, Ryman et al. (1981)did not simulate hunts featuring extreme selectionfor males. They found Ne/N ratios as low as 0.2 (i.e.,a population of 100 would experience genetic driftat the rate of an ideal population of 20), but point-ed out that extreme selectivity for males in the huntcould further reduce this ratio.

Harris and Allendorf (1989) varied huntingregimes for hypothetical grizzly bear (Ursus arctos)populations, finding relationships between the typeof hunt and Ne. In some cases, Ne/N increased fromthe nonhunted situation, because reproductive suc-

cess among males became more equitable.However, in hunting scenarios where the numberof males became limiting, Ne/N declined from itsunhunted level.

Wall (1989) examined demographics and het-erozygosity of white-tailed deer populations inTexas exposed to hunts with differing selectivities.Although he was unable to estimate variation inreproductive success, an important determinant inNe (Harris and Allendorf 1989), Wall (1989) com-pared “maximum Ne” among populations based onthe demographic parameters and assuming no dif-ferences in variance of reproductive success.Variable hunting strategies had profound effects onNe/N. However, because those hunting regimesreducing Ne/N often were designed to keep N high,Ne generally varied less than did Ne/N. For exam-ple, populations exposed to buck-only harvest hadlow Ne/N (because few bucks dominated breed-ing), but high Ne (because census population sizeremained high, being primarily a female popula-tion). In contrast, hunting regimes with a relativelyhigh female harvest (and more equitable sex ratios)had the highest Ne/N ratios, but lower Ne (becausethe total population was lower). His sampling ofgenetic attributes suggested, however, that geneticdrift (as documented by H) was a substantial con-cern only in the smallest, most isolated population.

Concerns about low Ne have been appropriatelyfocused on small or declining populations(Allendorf and Ryman 2002) rather than on thelarger populations typically subjected to sporthunting. Managers of sport hunts should be mind-ful, however, of the potential for undesirable genet-ic consequences of low Ne where high harvestrates produce severely skewed sex ratios. Sex ratiosof about 1 adult male:10 adult females were docu-mented for elk (Cervus elaphus, Leptich and Zager1991, Noyes et al. 1996) and mule deer (Scribner etal. 1991), and are probably common in otherspecies where males are selectively hunted. Weagree with Scribner et al. (1991) that large popula-tion size substantially reduces the concern aboutgenetic drift. Nevertheless, smaller breedinggroups of related individuals may occur within larg-er populations because of strong site fidelity byfemales (Scribner et al. 1991). Therefore, highlyskewed sex ratios may increase the frequency ofinbreeding even in the presence of little popula-tion-wide genetic drift. Most managers of ungulatepopulations attempt to prevent adult sex ratiosfrom reaching such extremes in order to maintainnormal breeding behavior. We believe that loss of

Commentary • Harris et al. 637

genetic variability, even if nested within a largerpopulation, is another reason to avoid highlyskewed sex ratios.

Hunting and deliberate selectionMany European wildlife managers have tradition-

ally attempted to alter antler or horn characteristicsof artiodactyls by selectively culling those consid-ered inferior (Webb 1960, Taber 1961, Hartl 1991,Sforzi and Lovari 2000). Culling of yearling white-tailed deer with poor antler development was alsosuggested in the southeast United States (Harmel1983, Cook 1984, Newsome 1984, but see Lukefahrand Jacobson 1998). Although such managementmight be seen as a partial correction to practiceswhere only the largest animals are taken, it is notwithout risks. By selecting for one particular traitof perceived value to humans, we believe it likelythat management simultaneously (if inadvertently)selects against other traits potentially of adaptivesignificance for the species (Voipio 1950, Klein etal. 1992). In particular, relatively rare alleles thatmight be important in a long-term evolutionary per-spective are vulnerable to loss when such selectionfor other traits takes place.

Research on red deer in Europe provided com-pelling evidence that deliberate selection couldhave unintended consequences. In France, Hartl etal. (1991, 1995) found that alleles at loci Idh-2, Me-1, and Acp-1 were associated with body and antlersize in red deer. Deliberate culling of yearling bullswith undesirable antler characteristics rapidlyincreased the frequency of an allele (Idh-2125) pos-itively correlated with number of antler points.Importantly, Pemberton et al. (1988) found thatjuvenile survival among juvenile female red deer inScotland heterozygous at this same Idh-2 locus washigher than among homozygous individuals. Thus,by selectively removing males with small antlersand thus reducing the frequency of the alternateallele at Idh-2, it appeared that French hunters mayhave also unwittingly selected for poor juvenile sur-vival (Hartl 1991). In general, when one phenotyp-ic trait is maximized other traits are inevitably (andprobably unknowingly) affected because life-histo-ry strategies inevitably involve trade-offs amongvarious fitness components related to demographicequilibrium (Pemberton et al. 1991). Thus, humanattempts to “improve” hunted species throughselective culling seem certain to produce unfore-seen consequences. Similarly, releasing penneddeer bred specifically for antler growth (Cook

1984) into the wild (to produce large “superbucks”) seems to us careless disregard for this fun-damental concept.

Hunting and unintentional selectionNorth American managers have often down-

played possible genetic consequence of selectivehunting, focusing instead on maximizing yieldeither of total animals or of trophy males. However,harvest regimes that are focused on removing largemales risk producing inadvertent directional selec-tion against the very characteristics (usually largeantlers or horns in artiodactyls) that hunters desire.We distinguished the effects that selective huntingmay have on genetic diversity generally (as indicat-ed by H) from those leading to loss of specific alle-les.

Selective hunting and H H is often (albeit not universally) thought to be

related to fitness in natural populations (Allendorfand Leary 1986, Britten 1996, Coltman et al. 1999).In white-tailed deer, studies have reported positivecorrelations between H and twinning (Johns et al.1977, Chesser and Smith 1987), fetal growth(Cothran et al. 1983), body weight (Smith et al.1982), body size (Chesser and Smith 1987), andantler size (Scribner et al. 1989). The last-namedauthors considered it likely that higher H resultedin higher metabolic efficiency, and thus decreasedmaintenance-energy requirements, leading to largerantlers. Fitzsimmons et al. (1995) found slightlygreater yearly horn growth among >6-year-oldbighorn sheep rams that were heterozygous at >2loci than among those heterozygous at a zero orone locus.

However, other studies have found no correla-tions between H and body mass or number ofantler points in white-tailed deer (Sheffield et al.1985) and red deer (Hartl et al. 1991). Chesser andSmith (1987) reported negative as well as positivecorrelations between H and components of repro-duction related to fitness. Further, the relationshipsinvolving antler or horn growth observed byScribner et al. (1989) and Fitzsimmons et al. (1995)occurred only for older age-classes. Because themajority of antler and horn growth occurred inyounger classes (for which no correlation with Hwas found), the total amount of variation attributa-ble to H class was low. Antler and horn growth alsoare known to respond to environmental factors(and the largest single determinant of size usually is

638 Wildlife Society Bulletin 2002, 30(2):634–643

age class), so it is difficult to distinguish the effectsof H on horn or antler size.

Can selection imposed by hunting reduce H?Fitzsimmons et al. (1995) voiced concerns thatselectively removing the largest rams by huntingwould, perhaps unintentionally, reduce geneticvariability in such populations. Although not specif-ically designed to examine such a possibility, thework of Wall (1989) provided some insight.Despite widely varying hunting regimes (and thusstanding age structures), he found no differences inH, as measured by allozymes from harvested deer,among populations examined. If antler qualitywere related to H, we might expect hunter-harvest-ed samples from those hunts featuring the greatestselectivity for trophies to exhibit higher H thanthose from less selective hunts.

We found that the evidence for selective removalof relatively more heterozygous individuals withinnatural populations was weak. Perhaps moreimportantly, loss of H caused by removal of individ-uals that tend to be heterozygous at specific loci isa reversible process. That is, even if heterozygoteswere selectively removed, heterozygous progenywould be regenerated the next generation by mat-ings between individuals that are homozygous fordifferent alleles (Mitton 1997). Thus, although pos-sible deleterious effects from selective removal ofheterozygous individuals bears monitoring, it doesnot appear to be a serious problem.

Selective hunting and changes in allelefrequency

A slightly different mechanism may come intoplay where hunter selectivity is based on pheno-typic traits such as horns or antlers. Such huntsmay unintentionally select against those very traitsby reducing the life span (and thus the reproduc-tive contribution) of individuals carrying specificalleles. Festa-Bianchet (2002) suggested that heavyhunting can alter selective pressures of femaleartiodactyls from those favoring high survival andlow maternal investment per litter to those favoringearly reproduction and lower survival. We sharewith Festa-Bianchet (2002) additional concernsabout long-term genetic consequences of trophyhunts on phenotypic characteristics of males.

The empirical literature is ambiguous onwhether hunting regimes focused on taking maleswith large horns or antlers unintentionally altersallele frequencies (as evidenced by phenotypicchanges). Dubas and Jezierski (1989) documenteddeclining antler quality and carcass weight by age

over a 6-year time period in European red deer,speculating that selective hunting may have playeda part. However, population density also increasedduring their study, confounding interpretation(Clutton-Brock et al. 1982). Ludwig and Hoefs(1995) discounted hunting as a possible factor intheir finding that Dall sheep (Ovis dalli) in a hunt-ed population had shorter horns than did those inthe adjacent (unhunted) Kluane National Parkdespite similar age distributions (horn circumfer-ences in the 2 populations did not differ). Solbergand Sæther (1994) reported no decline in antlersize over a 23-year period of moose harvest. In con-trast, Shea and Vanderhoof (1999) observed a reduc-tion in antler size of 2.5-year-old white-tailed deer 5years after initiation of a hunting regime intendedto increase antler size by prohibiting harvest ofsmall-antlered bucks. They attributed the unex-pected reduction to “high-grading” (i.e., selectivekilling) of bucks born earlier during their year ofbirth. Their data also showed that early birth wasassociated with larger antlers, leaving predominant-ly late-born bucks to survive to 2.5 years. Shea andVanderhoof (1999) evidently did not examine thegenetic basis of these changes, but selective hunt-ing of larger bucks may have changed frequenciesof alleles that contributed to antler size.

In most cases hunters prefer to harvest largeartiodactyls, and horn or antler size is generally cor-related with male fitness. However, it does not fol-low that hunting removes relatively more fit indi-viduals in all cases. Artiodactyl males begin withsmall horns or antlers that become progressivelylarger with age. Hunters selecting for individualswith the largest horns or antlers remove primarilyold individuals, not necessarily those with genomesconducive to producing large secondary sexualcharacteristics. Changes in allele frequenciescaused by selective hunting of large males may bebuffered by the genetic contributions of females,which will have most of the same alleles as malesbut are likely to be subject to differing selectivepressures. Finally, other factors may affect vulnera-bility to hunting independent of hunter selectivity.For example, DuFour et al. (1993) found that mal-lards in poorer body condition were more vulnera-ble to hunting than those in better condition.

Given these complexities and ambiguities, webelieve the simulation model constructed byThelen (1991; see also Hundertmark et al. 1993,1998) currently provides the best indication ofhow selective hunting might unintentionally alterthe genetic constitution of big game populations.

Commentary • Harris et al. 639

Thelen (1991) assumed that antler characteristics ofelk were polygenic traits, inherited in simpleMendelian fashion but with no dominance or epis-tasis. Antler size increased with age, but antler char-acteristics were also assumed to be an additive func-tion of multiple loci (i.e. the greater the number ofalleles favorable for large antlers, the larger theantlers). Heritability, the proportion of variation inantler size attributable to genotype, varied from 25%to 75%. Breeding success was controlled by bothage and antler size. Age-specific survival of bulls wasalso negatively correlated with antler size (i.e., a sur-vival cost of carrying heavy antlers was assumed).Hunting strategies were modeled to reflect variouspossible management objectives. Hunting regimesthat specified minimum antler sizes always reducedthe frequency of large antler alleles in modeled pop-ulations. When heritability of antler traits was mod-eled as 50%, allele frequencies were altered byapproximately 10–20% after 50 years. When heri-tability was assumed to be 75%, the 50-year reduc-tion in favorable allele frequency was on the orderof 20–25%; with heritability of 25%, the 50-yearreduction was about 10% (Figure 1).

By contrast, other hunting regimes had littleeffect on the frequency of existing alleles in thepopulation. A nonrestricted harvest strategy, in

which the sex and age of individuals harvested wasdirectly proportional to their abundance in thepopulation, had little effect on allele frequencyafter 50 years. The nonrestricted hunting regimealso resulted in relatively high yield overall, but lowyield of trophy males. But a split hunting regime, inwhich spike (mostly yearling) and >5 point (tro-phy) males were legally taken (but 2–4 point malesprotected) resulted simultaneously in moderateoverall harvest, moderate harvest of trophy males,and little change in large-antler alleles (Figure 1).

It is likely not coincidental the mortality patternproduced by the split hunting regime resembledthat of an ungulate population experiencing onlynonhunting mortality (high mortality in young ageclasses, low among mature animals, high again asanimals senesce). Thelen’s (1991) model suggestedthe possibility of achieving a sustainable harvest(including of trophy males) while avoiding sub-stantial alterations in allele frequencies, by moder-ating hunting pressure focused on large males andsimultaneously harvesting young males vulnerableto natural mortality. Klein et al. (1992) also recom-mended a split hunting regime as one that wouldproduce a good compromise between hunter satis-faction and long-term evolutionary concerns.

Thelen’s results were sensitive to heritability ofantler size (Figure 1). There is little doubt thatantler characteristics are heritable traits (Harmel1983, Williams et al. 1994, Lukefahr and Jacobson1998, Wang et al. 1999). However, all studies to datehave estimated heritability under captive condi-tions, thus the “unexplained” portion of antler vari-ability (i.e., that remaining after inheritance isexplained) has been low. In the wild, we wouldexpect age-specific antler size to vary considerablywith nutritional status as well as with genotype(Brown 1990).

An additional, important factor in such models isthe strength of reproductive advantage enjoyed bymales with desired traits. Conventional wisdomsuggests that males with the largest horns andantlers have the highest reproductive success(Solberg and Sæther 1994). But just how muchhigher? The rate of loss of alleles affecting horn andantler size on the population caused by hunterselection of large males would be higher if malesvulnerable to hunting dominate the breeding, andlower if smaller males (which would presumablysurvive a selective hunt) also make substantial con-tributions under natural conditions.

In unhunted red deer, paternity data showed notonly that dominant bulls had greater reproductive

Figure 1. Percent change after 50 years in frequency of allelesmodeled as favorable for large antler formation in elk, under 4alternative hunting strategies, and assuming antler heritabilityof 25%, 50%, and 75%. Under “nonrestricted” hunts, bullswere killed in the simulated hunt in proportion to their abun-dance; under “2-” and 4-point minimum” hunts, only bullswith these antler characteristics were killed; “spike and 5+points” hunts protected males with 2–4 points from huntingmortality. Adapted from Thelen (1991).

640 Wildlife Society Bulletin 2002, 30(2):634–643

success than did subordinates, but also that theirrelative success was even greater than had beenestimated from behavioral data (Pemberton et al.1992). Younger, smaller, or less dominant bulls didrelatively little breeding. In contrast, paternity dataon bighorn sheep in two intensively studied popu-lations showed that high-ranking rams, while stillmore successful than lower ranking rams, fatheredfewer lambs than would have been estimated fromonly observing their success at tending estrousewes (Hogg and Forbes 1997). Rams using “uncon-ventional” courting tactics associated with lowerrank were surprisingly effective in contributingtheir genes to subsequent generations. Thus, a vari-ety of alleles in bighorns may be transmitted to sub-sequent generations by smaller, younger rams thatwould be unaffected by strongly selective hunts.

We suspect that long-term changes in allele fre-quencies are a common attribute of terrestrial pop-ulations subjected to strongly selective hunting. Itis difficult to see how it could be otherwise, giventhat hunting often constitutes the largest source ofmortality (Festa-Bianchet, 2002). However, becauseage and environment exert major influences onsize, mating systems are often flexible, gene flowamong adjacent populations that vary in mortalitypatterns may replenish vulnerable alleles, heritabil-ities of phenotypic traits observable to hunters maybe low, and offtake rates usually are moderated inthe most strongly selective hunts, we expect suchchanges to occur gradually and to be undetectablefor many generation lengths.

Research directionsResearch into genetic effects of hunting has been

much less common than research into demograph-ic and behavioral effects of hunting. Several impor-tant questions remain unanswered: 1) under whatconditions is the alteration of gene flow amongpopulation subdivisions resulting from hunting ofsufficient magnitude to cause problems? 2) doesselective hunting preferentially take more het-erozygous individuals than would a random hunt,and if so, does this reduce heterozygosity in thepopulation? and 3) how much breeding is con-ducted by subordinate males, and how does thatchange under various hunting regimes? Answers tothese questions are likely to vary by species, andperhaps also geographically.

Now that genetic techniques allow paternitydetermination (e.g., Hogg and Forbes 1997, Croninet al. 1999), researchers are in a much better posi-

tion to understand male reproductive success, itscorrelates, and its variance. Understanding patternsof male reproductive success has implications forthe survival effects on other age classes of remov-ing dominant males, influences of removing domi-nant males on breeding activity of females, and ofthe effects of management on genetically effectivepopulation size (Ne), which, in turn, tells us aboutthe magnitude of genetic drift. It is most useful toadd paternity studies to populations already underintensive demographic study, but where these con-ditions exist, the laboratory expenses usually willbe justified by the insight gained. Where paternitystudies are conducted and phenotypic informationalso obtained, the opportunity exists to documentthe heritability of secondary male sexual character-istics of interest to hunters (usually horns orantlers) under wild conditions. The strength of her-itability remains a critical, but largely unknown,piece of the puzzle in considering the long-termevolutionary consequences of selective hunting inungulates (Rèale et al. 1999, Kruuk et al. 2000).

Indirect information on possible directionalselection of genomes can also be obtained fromexisting data sources (e.g., hunter check-stations)by careful examination of long-term data sets com-paring hunting intensity to trends in horn/antlersize by age. Confounding effects, such as popula-tion density and varying environmental conditions,will need to be carefully considered. Further labo-ratory analyses (e.g., using carcasses at huntercheck-stations) correlating the presence of specificalleles, as well as estimates of H, with male charac-teristics (coupled with sampling of unhunted indi-viduals) would further elucidate whether selectivehunting has disruptive effects on H and allelicdiversity.

ConclusionsWe began by expressing concern about the long-

term genetic consequences of hunting, but ourreview of the literature suggested little empiricalevidence of such consequences. We have hypothe-sized a number of characteristics of hunting andhunted species that may act to mitigate expectednegative effects. However, we found no grounds forcomplacency; studies designed to quantify geneticeffects have been rare, and the effects eliciting ourgreatest level of concern are subtle and difficult todetect without long-term monitoring.

We stress that demographic and genetic changesoccur on different time scales. Demographic

Commentary • Harris et al. 641

effects are often immediate and easily recognized;genetic changes occur over evolutionary timescales of many generations. Thus, although short-lived actions are unlikely to have genetic effectsover the long term, any genetic changes will be dif-ficult to detect because of the time scale overwhich they occur. For example, it would requirecareful study to detect the effects of a hunting man-agement scheme that decreased mean antler sizeby 4% per generation (i.e., 1% reduction/yr in aspecies where the mean age of reproduction was 4years). However, such a rate of change may havesubstantial effects over the long term. For example,this rate of change would reduce mean antler sizeby 30% in 50 years.

We urge managers to consider not only the main-tenance of genetic diversity, but also whether theprimary selective forces influencing adaptations inhunted populations have become artificial ratherthan natural. Where concern is justified, it is pru-dent to manage hunting such that the age-specificsurvival pattern (and thus age-specific breedingstructure) emulates that occurring in the absenceof hunting (Klein et al. 1992, Hundertmark et al.1993). Such hunting regimes will generally pro-duce little alteration in allele frequencies, have lowchance of causing extinction of rare alleles, mini-mize extremely skewed sex ratios (and thus haveless effect on Ne), and still allow for the huntingopportunities we cherish.

Acknowledgments. We thank P. Voipio for pro-viding a copy of his classic monograph on thegenetic effects of game management. M. S. Boyce,W. J. Sutherland, W. M. Ford, M. A. Cronin, J. T. Hogg,and 2 anonymous reviewers suggested improve-ments to earlier versions of this manuscript.Funding for R. B. Harris during initial portions ofthis work was provided by Safari ClubInternational. We thank J. Hensiek and P. Sandstromfor translation help, and the staff at the Maureenand Mike Mansfield Library at the University ofMontana for assistance.

Literature citedALLENDORF, F. W., AND R. F. LEARY. 1986. Heterozygosity and fit-

ness in natural populations of animals. Pages 57–76 in M.Soulé, editor. Conservation biology: the science of scarcityand diversity. Sinauer Associates, Sunderland, Massachusetts,USA.

ALLENDORF, F. W., AND N. RYMAN. 2002. The role of genetics inpopulation viability analysis. S. R. Beissinger, and D. R.McCullough, editors. Population viability analysis. University

of Chicago, Chicago, Illinois, USA. In press.BRITTEN, H. B. 1996. Meta-analysis of the association between

multilocus heterozygosity and fitness. Evolution 50:2158–2164.

BROWN, R. D. 1990. Nutrition and antler development. Pages426–441 in G. A. Bubenik, and A. B. Bubenik, editors. Horns,pronghorns, and antlers. Springer-Verlag, New York, NewYork, USA.

CHESSER, R. K., AND M. H. SMITH. 1987. Relationship of geneticvariation to growth and reproduction in the white-taileddeer. Pages 168–177 in C. M. Wemmer, editor. Biology andmanagement of the Cervidae. Smithsonian Institution,Washington, D.C., USA.

CLUTTON-BROCK, T. H., F. E. GUINNESS, AND S. D. ALBON. 1982. Reddeer: behavior and ecology of two sexes. University ofChicago, Chicago, Illinois, USA.

COLTMAN, D. W., J. G. PILKINGTON, J. A. SMITH, AND J. M. PEMBERTON.1999. Parasite-mediated selection against inbred Soay sheepin a free-living island population. Evolution 53: 1259–1267.

COTHRAN, E. G., R. CHESSER, M. H. SMITH, AND P. E. JOHNS. 1983.Influences of genetic variability and maternal factors on fetalgrowth in white-tailed deer. Evolution 37:282–291.

COOK, R. L. 1984. Texas. Pages 457–474 in L. K. Halls, editor.White-tailed deer: ecology and management. Stackpole,Harrisburg, Pennsylvania, and Wildlife Management Institute,Washington, D.C., USA.

CRONIN, M. A., M. E. NELSON, AND D. F. PAC. 1991. Spatial hetero-geneity of mitochondrial DNA and allozymes among popula-tions of white-tailed deer and mule deer. Journal of Heredity82:118–127.

CRONIN, M. A., R. SHIDELER, J. HECHTEL, C. STROBECK, AND D.PAETKAU. 1999. Genetic relationships of grizzly bears (Ursusarctos) in the Prudhoe Bay Region of Alaska: inference frommicrosatellite DNA, mitochondrial DNA, and field observa-tions. Journal of Heredity 90:622–628.

DUBAS, J. W. AND W. JEZIERSKI. 1989. Effekte der jagdlichenAuslese beim männlichen Rotwild. (Effects of selective hunt-ing on male red deer). Zeitschrift Für Jagdwissenschaft 35:192–197. In German.

DUFOUR, K. W., C. D. ANKNEY, AND P. J. WEATHERHEAD. 1993.Condition and vulnerability to hunting among mallards stag-ing at Lake St. Clair, Ontario. Journal of Wildlife Management57:209–215.

ELLSWORTH, D. L., R. L. HONEYCUTT, N. J. SILVY, M. H. SMITH, J. W.BICKHAM, AND W. D. KLIMSTRA. 1994. White-tailed deer restora-tion to the southeastern United States: evaluating geneticvariation. Journal of Wildlife Management 58:686–697.

FESTA-BIANCHET, M. 2002. Exploitative wildlife management as aselective pressure for life-history evolution of large mammals:in M. Festa-Bianchet and M. Apollonio, editors. Wildlife behav-ior and conservation. Island, Washington, D.C. USA. In press.

FITZSIMMONS, N. N., S. W. BUSKIRK, AND M. H. SMITH. 1995.Population history, genetic variability, and horn growth inbighorn sheep. Conservation Biology 9:314–323.

FITZSIMMONS, N. N., S. W. BUSKIRK, AND M. H. SMITH. 1997. Geneticchanges in reintroduced Rocky Mountain bighorn sheeppopulations. Journal of Wildlife Management 61:863–872.

FRATI, F., G. B. HARTL, AND S. LOVARI. 2000. Does protection fromhunting favour genetic uniformity in the red fox? Zeitschriftfür Säugetierkunde 65:76–83.

FRATI, F., G. B. HARTL, S. LOVARI, M. DELIBES, AND G. MARKOV. 1998.Quaternary radiation and genetic structure of the red foxVulpes vulpes in the Mediterranean Basin, as revealed by

642 Wildlife Society Bulletin 2002, 30(2):634–643

allozymes and mitochondrial DNA. Journal of Zoology 245:43–51.

GINSBERG, J. R., AND E. J. MILNER-GULLAND. 1994. Sex-biased har-vesting and population dynamics in ungulates: implicationsfor conservation and sustainable use. Conservation Biology8:157–166.

HARMEL, D. E. 1983. Effects of genetics on antler quality andbody size in white-tailed deer. Pages 339–348 in R. D. Brown,editor. Antler development in Cervidae. Caesar KlebergWildlife Research Institute, Kingsville, Texas, USA.

HARRIS, R. B. AND F. W. ALLENDORF. 1989. Genetically effectivepopulation size of large mammals: an assessment of estima-tors. Conservation Biology 3:181–191.

HARTL, G. B. 1991. The influence of game management on allel-ic variation in large mammals of central Europe. Pages95–108 in E. Randi and M. Spagnesi, editors. Genetics andWildlife Conservation, Supplemento alle Richerche diBiologia della Selvaggina. Volume XVIII. Bologna, Italy.

HARTL, G. B., F. KLEIN, R. WILLING, M. APOLLONIO, AND G. LANG.1995. Allozymes and the genetics of antler development inred deer (Cervus elaphus). Journal of Zoology. 237:83–100.

HARTL, G. B., G. LANG, F. KLEIN, AND R. WILLING. 1991.Relationships between allozymes, heterozygosity and mor-phological characters in red deer (Cervus elaphus), and theinfluence of selective hunting on allele frequency distribu-tions. Heredity 66:343–350.

HOGG, J. T. 2000. Mating systems and conservation at large spa-tial scales. Pages 214–252 in M. Appolonia and M. Festa-Bianchet, editors. Vertebrate mating systems. WorldScientific, London, United Kingdom.

HOGG, J. T., AND S. H. FORBES. 1997. Mating in bighorn sheep: fre-quent male reproduction via a high-risk “unconventional”tactic. Behavioral Ecology and Sociobiology 41:33–48.

HUNDERTMARK, K. J., T. H. THELEN, AND R. T. BOWYER. 1998.Effects of population density and selective harvest on antlerphenotype in simulated moose populations. Alces 34:375–383.

HUNDERTMARK, K. J., T. H. THELEN, AND C. C. SCHWARTZ. 1993.Population and genetic effects of selective harvest strategiesin moose: a modeling approach. Alces 29:225–234.

JOHNS, P. E., R. BACCUS, M. N. MANLOVE, J. E. PINDER, AND M. H.SMITH. 1977. Reproductive patterns, productivity and genet-ic variability in adjacent white-tailed deer populations.Proceedings of the Annual Conference SoutheasternAssociation Fish Wildlife Agencies 31:167–172.

KLEIN F., G. B. HARTL, A. SCHREIBER, AND G. LANG. 1992. The biological management of red deer (Cervus elaphus L.).Pages 113–115 in F. Spitz, G. Janeau, G. Gonzalez, and S.Aulagnier, editors. Ongules/Ungulates 91. Proceedings of theInternational Symposium, 2–6 September, 1991, Toulouse,France.

KURT, F., G. B. HARTL, AND F. VOELK. 1993. Breeding strategies andgenetic variation in European roe deer Capreolus capreoluspopulations. Acta Theriologica 38 (Supplement 2):187–194.

KRUUK, L. E. B., T. H. CLUTTON-BROCK, J. SLATE, J. M. PEMBERTON, S.BROTHERSTONE, AND F. E. GUINNESS. 2000. Heritability of fitnessin a wild mammal population. Proceedings of the NationalAcademy of Sciences 97:698–703.

LAW, R. 2000. Fishing, selection, and phenotypic evolution. ICESJournal of Marine Science 57:659–668.

LEE, T. E. JR., J. N. DERR, J. W. BICKHAM, AND T. L. CLARK. 1989.Genetic variation in pronghorn from west Texas. Journal ofWildlife Management 53:890–896.

LEPTICH D. J. AND P. ZAGER. 1991. Road access managementeffects on elk mortality and population dynamics. Pages126–131 in A. G. Christensen, L. J. Lyon, and T. N. Lonner, edi-tors. Proceedings of a Symposium on Elk Vulnerability.Montana State University, 10–12 April, 1991, Bozeman, USA.

LITTLE, R. M., T. M. CROWE, AND GRANT, W. S. 1993. Does huntingaffect the demography and genetic structure of the greywingfrancolin Francolinus africanus? Biodiversity andConservation 2:567–585.

LUDWIG, J., AND M. HOEFS. 1995. Eine populationsgenetischeStudie an Dallschafen (Ovis dalli dalli) im südwestlichenYukongebiet Kanadas (A population genetic study of Dallsheep in southwestern Yukon Territory, Canada). ZeitschriftFür Jagdwissenschaft 41:256–266. In German.

LUKEFAHR, S. D., AND H. A. JACOBSON. 1998. Variance componentanalysis and heritability of antler traits in white-tailed deer.Journal of Wildlife Management 62:262–268.

MATHEWS, N. E., AND W. F. PORTER. 1993. Effects of social struc-ture on genetic structure of free-ranging white-tailed deer inthe Adirondack Mountains. Journal of Mammalogy 74:33–43.

MITTON, J. B. 1997. Selection in natural populations. OxfordUniversity, New York, New York, USA.

NEDBAL, M. A., R. L. HONEYCUTT, S. G. EVANS, R. M. WHITING JR., AND

D. R. DIETZ. 1997. Northern bobwhite restocking in eastTexas: a genetic assessment. Journal of Wildlife Management61:854–863

NEWSOME, J. D. 1984. Coastal plain. Pages 367–380 in L. K. Halls,editor. White-tailed deer: ecology and management.Stackpole, Harrisburg, Pennsylvania, and WildlifeManagement Institute, Washington, D.C., USA.

NOYES, J. H., B. K. JOHNSON, L. D. BRYANT, S. L. FINDHOLT, AND J. W.THOMAS. 1996. Effects of bull age on conception dates andpregnancy rates of cow elk. Journal of Wildlife Management60:508–517.

PEMBERTON, J. M., S. D. ALBON, F. E. GUINNESS, AND T. H. CLUTTON-BROCK. 1991. Countervailing selection in different fitnesscomponents in female red deer. Evolution 45:93–103.

PEMBERTON, J. M., S. D. ALBON, F. E. GUINNESS, T. H. CLUTTON-BROCK,AND G. A. DOVER. 1992. Behavioral estimates of male matingsuccess tested by DNA fingerprinting in a polygynous mam-mal. Behavioral Ecology 3:66–75.

PEMBERTON, J. M, S. D. ALBON, F. E. GUINNESS, T. H. CLUTTON-BROCK,AND R. J. BERRY. 1988. Genetic variation and juvenile survivalin red deer. Evolution 42:921–934.

PETIT, E., S. AULAGNIER, R. BON, M. DUBOIS, AND R. B. CROUAU. 1997.Genetic structure of populations of the Mediterranean mou-flon (Ovis gmelini). Journal of Mammalogy 78:459–467.

PURDUE, J. R., M. H. SMITH, AND J. C. PATTON. 2000. Femalephilopatry and extreme spatial genetic heterogeneity inwhite-tailed deer. Journal of Mammalogy 81:179–185.

RÈALE, D., M. FESTA-BIANCHET, AND J. T. JORGENSON. 1999.Heritability of body mass varies with age and season in wildbighorn sheep. Heredity 83:526–532.

RHODES, O. E., JR. AND M. H. SMITH. 1992. Genetic perspectivesin wildlife management: the case of large herbivores. Pages985–996 in D. R. McCullough, and R. H. Barrett, editors.Wildlife 2001: populations. Elsevier Science, London, UnitedKingdom.

ROBERTSON, P.A., D. R. WISE, AND K. A. BLAKE. 1993. Flying abilityof different pheasant strains. Journal of Wildlife Management57:778–782.

RYMAN, N., R. BACCUS, C. REUTERWALL, AND M. H. SMITH. 1981.Effective population size, generation interval, and potential

Commentary • Harris et al. 643

loss of genetic variability in game species under differenthunting regimes. Oikos 36:257–266.

SCRIBNER, K. T., M. H. SMITH, AND R. K. CHESSER. 1997. Spatial andtemporal variability of microgeographic genetic structure inwhite-tailed deer. Journal of Mammalogy 78:744–755.

SCRIBNER, K. T., M. H. SMITH, R. A. GARROTT, AND L. H. CARPENTER.1991. Temporal, spatial, and age-specific changes in genotyp-ic composition of mule deer. Journal of Mammalogy 72:126–137.

SCRIBNER, K. T., M. H. SMITH, AND P. E. JOHNS. 1989. Environ-mental and genetic components of antler growth in white-tailed deer. Journal of Mammalogy 70:284–291.

SFORZI, A., AND S. LOVARI. 2000. Some effects of hunting on wildmammalian populations. Ibex Journal of Mountain Ecology5:165–171.

SHEA, S. M., AND R. E. VANDERHOOF. 1999. Evaluation of a five-inchregulation for increasing antler size of harvested deer innorthwest Florida. Proceedings of the Annual SoutheasternDeer Study Group 22:18–19.

SHEFFIELD, S. R., R. P. MORGAN, II., G. A. FELDHAMMER, AND D. M.HARMAN. 1985. Genetic variation in white-tailed deer(Odocoileus virginianus) populations in western Maryland.Journal of Mammalogy 66:243–255.

SMITH, M. H., R. K. CHESSER, E. G. COTHRAN, AND P. E. JOHNS. 1982.Genetic variability and antler growth in a natural populationof white-tailed deer. Pages 365–387 in R. D. Brown, editor.Antler development in Cervidae. Caesar Kleberg WildlifeResearch Institute, Kingsville, Texas, USA.

SOLBERG, E. J., AND B.-E. SÆTHER. 1994. Male traits as life-historyvariables: annual variation in body mass and antler size inmoose (Alces alces). Journal of Mammalogy 75:1069–1079.

SOLBERG, E. J., A. LOISON, B.-E. SÆTHER, AND O. STRAND. 2000. Age-specific harvest mortality in a Norwegian moose Alces alcespopulation. Wildlife Biology 6:41–52.

STOKES, T. K., J. M. MCGLADE, AND R. LAW. 1993. The exploitationof evolving resources. Springer-Verlag, Berlin, Germany.

STRANDGAARD, H., AND V. SIMONSEN. 1993. Genetic differentiationin populations of red deer, Cervus elaphus, in Denmark.Hereditas 119:171–177.

SUTHERLAND, W. J. 1990. Evolution and fisheries. Nature 344:814–815

TABER, R. D. 1961. Wildlife administration and harvest in Poland.Journal of Wildlife Management 25:353–363

THELEN, T. H. 1991. Effects of harvest on antlers of simulatedpopulations of elk. Journal of Wildlife Management 55:243–249

VOIPIO, P. 1950. Evolution at the population level with specialreference to game animals and practical game management.Papers on Game Research 5. Finnish Foundation for GamePreservation. Helsinki, Finland.

WALL, A. W. 1989. Demographic and genetic variation in differ-entially managed deer herds. Dissertation, Stephen F. AustinState University, Nacogdoches, Texas, USA.

WANG, Z., R. C. YANG, L. A. GOONEWARDENE, AND C. HUEDEPOHL.1999. Genetic analysis of velvet antler yield in farmed elk(Cervus elaphus). Canadian Journal of Animal Science 79:569–571.

WARNER, R. E., J. B. KOPPELMAN, AND D. P. PHILIPP. 1988. A bio-chemical genetic evaluation of ring-necked pheasants.Journal of Wildlife Management 52:108–109.

WEBB, W. L. 1960. Forest wildlife management in Germany.Journal of Wildlife Management 24:147–161.

WILLIAMS, J. D., W. F. KRUEGER, AND D. H. HARMEL. 1994.Heritabilities for antler characteristics and body weight inyearling white-tailed deer. Heredity 73:78–83.

WRIGHT, S. 1969. Evolution and the genetics of populations.Volume 2. The theory of gene frequencies. University ofChicago, Chicago, Illinois, USA.

Address for Richard B. Harris: Wildlife Biology Program, Schoolof Forestry, University of Montana, Missoula, MT 59812, USA;e-mail: rharris@montana.com. Address for William A. Wall:Safari Club International, 441-E Carlisle Dr., Herndon, VA20170, USA. Address for Fred W. Allendorf: Division ofBiological Sciences, University of Montana, Missoula, MT59812, USA.

Richard B. (Rich) Harris (photo) is research associate in theWildlife Biology Program at the University of Montana, and alsoserves as editor of the journal Ursus. He received his Ph.D.from the University of Montana and later worked as program-matic biologist for the Montana Department of NaturalResources and Conservation. His research interests includeenhancing incentives for local people to conserve wildlife inwestern China, where he has conducted surveys and workedwith hunting programs since 1988. William A. (Bill) Wall cur-rently is senior scientist for wildlife conservation for Safari ClubInternational Foundation. He received his Ph.D. from StephenF. Austin State University in Nacogdoches, TX. He currently isdeveloping conservation hunting programs for argali sheep inCentral Asia. Fred W. Allendorf is professor of biology at theUniversity of Montana. His primary research interest in theapplication of genetic principles to problems in the conserva-tion and management of natural populations.