Sperm competition and diversity in rodent copulatory behaviour

The copulatory behaviour of male mammals is characterized by striking diversity in patterns of copulatory stimulation and ejaculation frequency. We conducted comparative analyses of rodents to investigate the potential influence of sperm competition in the evolution of copulatory behaviour. We found that increasing sperm competition is associated with faster rates of stimulation and earlier ejaculation among species in which males perform multiple intromissions prior to ejaculation, but with no overall change in the number of intromissions per ejaculation. Increasing sperm competition is also associated with a higher frequency of repeated ejaculations with the same female, and with shorter refractory periods between repeated copulations. Increasing sperm competition level thus appears to select for male ability to ejaculate more rapidly and more frequently with each female mated. As prolonged copulations are known to reduce female remating rates, these findings indicate that males may experience opposing selection pressures on copulatory behaviour with respect to offensive and defensive aspects of sperm competition. We conclude that sperm competition is likely to be an important selection pressure explaining diversity in mammalian copulatory behaviour.     

Frequent Failure of Male Monopolization Strategies as a Cost of Female Choice in the Black Widow Spider Latrodectus tredecimguttatus

In many mating systems, males strive for securing paternity through monopolizing females. As male monopolization attempts often contradict female interests, this conflict may fuel an evolutionary arms race. In the widow spider genus Latrodectus, females are commonly polyandrous, whereas males are monogynous, hence restricted to mate with a single female, making paternity protection particularly important. Potential mating plugs (specialized embolus sclerites of male copulatory organs) have been discovered in the complex female genital tracts of several Latrodectus species. In this study, we investigated mating strategies in the Mediterranean black widow spider Latrodectus tredecimguttatus and tested the adaptive value of female attacks against male monopolization efforts. In a double mating experiment, we manipulated the number of insertions (=copulations) for first and second males to assess female behaviour and male embolus sclerite placement success. Our results indicate that first males′ embolus sclerites inside the females′ sperm stores physically block sclerites of subsequent males. While female attacks did not affect the deposition of potential mating plugs, they significantly reduced copulation duration. Irrespective of female aggression, male sclerite placement failure occurred frequently, but large males were more successful than smaller competitors. We suggest that the complex genital morphology in both sexes may have co‐evolved antagonistically and female morphology could serve to favour large males for fertilization.     

Paternity analyses in wild‐caught and laboratory‐reared Caribbean cricket females reveal the influence of mating environment on post‐copulatory sexual selection

Polyandry is ubiquitous in insects and provides the conditions necessary for male‐ and female‐driven forms of post‐copulatory sexual selection to arise. Populations of Amphiacusta sanctaecrucis exhibit significant divergence in portions of the male genitalia that are inserted directly into the female reproductive tract, suggesting that males may exercise some post‐copulatory control over fertilization success. We examine the potential for male–male and male–female post‐copulatory interactions to influence paternity in wild‐caught females of A. sanctaecrucis and contrast our findings with those obtained from females reared in a high‐density laboratory environment. We find that female A. sanctaecrucis exercise control by mating multiple times (females mount males), but that male–male post‐copulatory interactions may influence paternity success. Moreover, post‐copulatory interactions that affect reproductive success of males are not independent of mating environment: clutches of wild‐caught females exhibit higher sire diversity and lower paternity skew than clutches of laboratory‐reared females. There was no strong evidence for last male precedence in either case. Most attempts at disentangling the contributions of male–male and male–female interactions towards post‐copulatory sexual selection have been undertaken in a laboratory setting and may not capture the full context in which they take place – such as the relationship between premating and post‐mating interactions. Our results reinforce the importance of designing studies that can capture the multifaceted nature of sexual selection for elucidating the role of post‐copulatory sexual selection in driving the evolution of male and female reproductive traits, especially when different components (e.g. precopulatory and post‐copulatory interactions) do not exert independent effects on reproductive outcomes.     

Female Remating Behavior in a Lekking Moth

In polyandrous species, male reproductive success will at least partly be determined by males' success in sperm competition. To understand the potential for post‐mating sexual selection, it is therefore important to assess the extent of female remating. In the lekking moth Achroia grisella, male mating success is strongly determined by female choice based on the attractiveness of male ultrasonic songs. Although observations have indicated that some females will remate, only little is known about the level of sperm competition. In many species, females are more likely to remate if their first mating involved an already mated male than if the first male was virgin. Potentially, this is because mated males are less well able to provide an adequate sperm supply, nutrients, or substances inhibiting female remating. This phenomenon will effectively reduce the strength of pre‐copulatory sexual selection because attractive males with high mating success will be more susceptible to sperm competition. We therefore performed an experiment designed both to provide a more precise estimate of female remating probability and simultaneously to test the hypothesis that female remating is influenced by male mating history. Overall, approximately one of five females remated with a second male. Yet, although females mated to non‐virgin males were somewhat more prone to remate, the effect of male mating history was not significant. The results revealed, however, that heavier females were more likely to remate. Furthermore, we found that females' second copulations were longer, suggesting that, in accordance with theory, males may invest more sperm in situations with an elevated risk of sperm competition.     

The copulatory plug delays ejaculation by rival males and affects sperm competition outcome in house mice

Females of many species mate with multiple males (polyandry), resulting in male–male competition extending to post‐copulation (sperm competition). Males adapt to such post‐copulatory sexual selection by altering features of their ejaculate that increase its competitiveness and/or by decreasing the risk of sperm competition through female manipulation or interference with rival male behaviour. At ejaculation, males of many species deposit copulatory plugs, which are commonly interpreted as a male adaptation to post‐copulatory competition and are thought to reduce or delay female remating. Here, we used a vertebrate model species, the house mouse, to study the consequences of copulatory plugs for post‐copulatory competition. We experimentally manipulated plugs after a female's first mating and investigated the consequences for rival male behaviour and paternity outcome. We found that even intact copulatory plugs were ineffective at preventing female remating, but that plugs influenced the rival male copulatory behaviour. Rivals facing intact copulatory plugs performed more but shorter copulations and ejaculated later than when the plug had been fully or partially removed. This suggests that the copulatory plug represents a considerable physical barrier to rival males. The paternity share of first males increased with a longer delay between the first and second males' ejaculations, indicative of fitness consequences of copulatory plugs. However, when males provided little copulatory stimulation, the incidence of pregnancy failure increased, representing a potential benefit of intense and repeated copulation besides plug removal. We discuss the potential mechanisms of how plugs influence sperm competition outcome and consequences for male copulatory behaviour.     

Sperm competition risk and male genital anatomy: comparative evidence for reduced duration of female sexual receptivity in primates with penile spines

Selection pressures influencing the way in which males stimulate females during copulation are not well understood. In mammals, copulatory stimulation can influence female remating behaviour, both via neuroendocrine mechanisms mediating control of sexual behaviour, and potentially also via effects of minor injury to the female genital tract. Male adaptations to increase copulatory stimulation may therefore function to reduce sperm competition risk by reducing the probability that females will remate. This hypothesis was tested using data for primates to explore relationships between male penile anatomy and the duration of female sexual receptivity. It was predicted that penile spines or relatively large bacula might function to increase copulatory stimulation and hence to reduce the duration of female sexual receptivity. Results of the comparative analyses presented show that, after control for phylogenetic effects, relatively high penile spinosity of male primates is associated with a relatively short duration of female sexual receptivity within the ovarian cycle, although no evidence was found for a similar relationship between baculum length and duration of female sexual receptivity. The findings presented suggest a new potential function for mammalian penile spines in the context of sexual selection, and add to growing evidence that sperm competition and associated sexual conflict are important selection pressures in the evolution of animal genitalia.     

Broken Copulatory Organs are Low-Cost Adaptations to Sperm Competition in Redback Spiders

In some spiders, a discrete portion of the male's copulatory organ (the apical sclerite) breaks off during copulation and remains in the female's reproductive tract. Apical sclerites may prevent insemination by rivals (sperm competition), stimulate females to favourably bias paternity (cryptic choice) or breakage may reflect sexual conflict over copulation duration with little direct effect on paternity. It has been assumed that any benefits of organ breakage are balanced by a large cost (male sterility) in species where males could otherwise mate multiply, but this has never been experimentally tested. We examined these ideas in the Australian redback spider (Latrodectus hasselti Thorell 1870, Araneae: Theridiidae), a species where males are functionally sterile after one normal mating. We experimentally removed sclerites and found males were able to mate, had similar copulation durations and transferred similar numbers of sperm as males with intact sclerites. Benefits of organ breakage were examined by forcing intact, rival males to inseminate the same or opposite reproductive tracts (female have paired, independent tracts in this taxon) and assessing paternity as a function of sclerite location. As predicted, apical sclerites were typically deposited at the entrance to the female's sperm storage organ, where they could physically block insemination by rivals. First male precedence was common when males inseminated the same tract and deposited sclerites at the entrance to the spermatheca, but not when sclerites were found elsewhere in the tract, or when rivals inseminated opposite tracts (where physically blocking rivals was impossible). Our data show that, in redbacks, copulatory organ breakage is not a side‐effect of sexual conflict, is unlikely to be a cue for cryptic female choice, but allows males to avoid sperm competition. Moreover, copulatory organ damage can have minimal reproductive cost for males, so assumptions of sterility after organ breakage are unjustified without supporting data.     

Copulatory behaviour of the bank voleMyodes glareolus: matings with one and two males do not make a difference

In promiscuous species in which females mate with more than one male during oestrus, males may increase their sperm expenditure or change their copulatory behaviour in response to the risk of sperm competition. I used an experimental approach to investigate the pattern of copulatory behaviour of the bank voleMyodes glareolus Schreber, 1780 depending on whether the female mated with one or two males. The work showed that the copulatory period of the bank vole lasted about 80 minutes and consisted of 4–5 ejaculatory series, with multiple intromissions preceding ejaculation. There were no significant changes in number of intromissions across the first four ejaculatory series, but I did find a relationship between number of intromissions and first ejaculation latency; also, ejaculation latencies grew shorter as the ejaculatory series proceeded. Litter size did not differ significantly between females that mated with one male and those mating with two, nor did the reproductive success of males that mated with the same female. Mating with an oestrus female appears to be advantageous for bank vole males even if they mate as the second one, and the risk of sperm competition did not trigger changes in male copulatory behaviour. The similar durations of the copulatory period and patterns of change of ejaculation latencies during copulations with one and two males point to the role of the female in temporal copulatory behaviour of the bank vole.     

A paternity advantage for speedy males? Sperm precedence patterns and female re-mating frequencies in a sexually cannibalistic praying mantid

Scramble competition polygyny is expected when females and/or resources are widely dispersed and not easily monopolized by males, or when there is an abundance of mates during an extremely restricted reproductive period. Additional factors such as first male sperm precedence or low female re-mating rate might further explain the propensity of males to engage in scramble competition. The sexually cannibalistic praying mantid Pseudomantis albofimbriata exhibits a polygynous mating system, where females exist in low-density populations and male competition manifests as the race to find females rather than as direct physical fighting. Here, we aim to determine whether there is a paternity advantage for the first-male to mate and/or a low frequency of female re-mating. First, we determined sperm precedence patterns in P. albofimbriata using the sterile male technique. Second, we tested the likelihood of female re-mating in P. albofimbriata by comparing the close-range approach behaviour and frequency of successful mating attempts for males when paired with virgin as opposed to recently mated females, and by comparing the frequency of long-distance male attraction between virgin and mated females. We found no paternity advantage for the first male to mate, rather a second male advantage. Although mated females were not rejected by males when approached from close-range, they were chemically unattractive to males searching from a distance. Since initial mate attraction in many praying mantids, including P. albofimbriata, is mediated via long-distance chemical communication, we believe the latter result is more ecologically relevant and therefore more important. These results suggest that the relatively low frequency of female re-mating observed in P. albofimbriata may be an additional factor driving scramble competition in this system.     

The combined effects of pre‐ and post‐copulatory processes are masking sexual conflict over mating rate in Gerris buenoi

In polygynandrous animals, post‐copulatory processes likely interfere with precopulatory sexual selection. In water striders, sexual conflict over mating rate and post‐copulatory processes are well documented, but their combined effect on reproductive success has seldom been investigated. We combine genetic parentage analyses and behavioural observations conducted in a competitive reproductive environment to investigate how pre‐ and post‐copulatory processes influence reproductive success in Gerris buenoi Kirkaldy. Precopulatory struggles had antagonistic effects on male and female reproductive success: efficiently gaining copulations was beneficial for males, whereas efficiently avoiding copulations was profitable for females. Also, high mating rates and an intermediate optimal resistance level of females supported the hypothesis of convenience polyandry. Contrary to formal predictions, high mating rates (i.e. the number of copulations) did not increase reproductive success in males or decrease reproductive success in females. Instead, the reproductive success of both sexes was higher when offspring were produced with several partners and when there were few unnecessary matings. Thus, male and female G. buenoi displayed different interests in reproduction, but post‐copulatory processes were masking the effects of copulatory mating success on reproductive success. Given the high mating rates observed, sperm competition could easily counter the effect of mating rates, perhaps in interaction with cryptic female choice and/or fecundity selection. Our study presents a complex but realistic overview of sexual selection forces at work in a model organism for the study of sexual conflict, confirming that insights are gained from investigating all episodes in the reproduction cycle of polygynandrous animals.     

Rival male relatedness does not affect ejaculate allocation as predicted by sperm competition theory

When females are sexually promiscuous, the intensity of sperm competition for males depends on how many partners females mate with. To maximize fitness, males should adjust their copulatory investment in relation to this intensity. However, fitness costs associated with sperm competition may not only depend on how many males a female has mated with, but also how related rival males are. According to theoretical predictions, males should adjust their copulatory investment in response to the relatedness of their male rival, and transfer more sperm to females that have first mated with a non-sibling male than females that have mated to a related male. Here, for the first time, we empirically test this theory using the Australian field cricket Teleogryllus oceanicus. We expose male crickets to sperm competition from either a full sibling or non-sibling male, by using both the presence of a rival male and the rival male's actual competing ejaculate as cues. Contrary to predictions, we find that males do not adjust ejaculates in response to the relatedness of their male rival. Instead, males with both full-sibling and non-sibling rivals allocate sperm of similar quality to females. This lack of kin biased behaviour is independent of any potentially confounding effect of strong competition between close relatives; kin biased behaviour was absent irrespective of whether males were raised in full sibling or mixed relatedness groups.     

Influence of male courtship intensity and male–male competition on paternity distribution in Hermann's tortoise,Testudo hermanni hermanni (Chelonia: Testudinidae)

Honest‐advertisement models of sexual selection suggest that condition‐dependent male secondary sexual characters could function as reliable signals of male quality, enabling females to discriminate among potential partners, both in the pre‐ and post‐copulatory phases. In this context, many studies have revealed the importance of promiscuous mating systems and female sperm storage in determining the occurrence of such a model of sexual selection. By contrast, few studies have investigated the presence and extent of post‐copulatory female choice in chelonian species. The present study aimed to investigate the effect of male size, male–male competition, and courtship intensity on paternity distribution in Testudo hermanni hermanni, combining behavioural and genetic data. We created experimental groups composed of two males of different sizes and three or four randomly selected females. Observations conducted during social interactions between males revealed that a hierarchy, unrelated to male size, was soon established: Alpha males were more aggressive towards competitors and courted females more intensively. Alpha males also achieved a higher mounting success than Beta males. Paternity analysis performed on hatchlings produced from experimental females revealed that male reproductive success was not correlated with male–female size ratio. Finally, despite the higher mounting success of Alpha males, paternity analysis revealed that male reproductive success did not differ between Alpha and Beta males. © 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 111 , 656–667.     

Sperm Competition and Copulation Intervals of the White Spoonbill (Platalea leucorodia,Aves, Threskiornithidae)

Some aspects of sperm competition were studied in the white spoonbill (Platalea leucorodia) breeding in Doñana National Park (SW Spain). Shorter pair copulation intervals occurred during the prelaying period, when females were subjected to a relatively high frequency of extra-pair copulations. Pair copulation intervals with an intermediate extra-pair copulation by the male mate were longer than those without extra-pair copulation. This result indicates that males need a time of recovery between copulations before they can perform another. Extra-pair copulations by the females did not affect the length of intervals between pair copulations. There were no differences between the lengths of the intervals between an extra-pair copulation by the female and the following pair copulation for cases in which the male mate detected an intruder male attempting copulation with his mate and those in which the intruder remained undetected. However, the correlations obtained between copulatory intervals for detected and undetected cases suggest a copulatory response by their mates, although affected by the required recovery time between copulations by the males. Finally, since extra-pair copulations mainly occurred while male mates were collecting nest material, they engaged in this activity shortly after pair copulations, probably to avoid a last-male advantage under the sperm competition pressure.     

Copulatory wounds in the monandrous ant species <Emphasis Type="Italic">Formica japonica</Emphasis> (Hymenoptera,Formicidae)

Males of several insect species inflict wounds on female genitalia during copulation, but the significance of such copulatory wounds for males is not clear. I compared the genitalia of virgin and mated Formica japonica females and for the first time report the occurrence of copulatory wounds in this monandrous ant species. All inseminated females examined had two types of melanized patches, indicating wound repair, and the serrated penis valves and sharp-pointed volsellar digitus of male genitalia were the likely instruments of these wounds. Physically damaging mating in monandrous species supports the view that copulatory wounds do not necessarily contribute to postcopulatory fitness gains for males via advantages in sperm competition or cryptic female choice. Received 10 September 2007; revised 15 October 2007; accepted 16 October 2007.     

Copulatory Plug Displacement Evidences Sperm Competition in Lemur catta

Male displacement of copulatory (sperm) plugs from female vaginas provides further evidence for sperm competition in ring-tailed lemurs (Lemur catta), a gregarious prosimian species with a multimale, multifemale mating system. During two mating seasons, I studied two groups of free-ranging ring-tailed lemurs on St. Catherines Island, GA, USA. I observed 22 mating pairs in which males achieved penile intromission. Copulatory plug displacement by males occurred in 9 cases. Plugs were displaced during copulation by male penes upon withdrawl following deep vaginal thrusting. In every case of copulatory plug displacement, the male displacing a plug mated to ejaculation with the estrous female. In a mating system in which females typically mate with more than one male during estrous, often in succession, copulatory plug displacement may function to disrupt or preclude other males' successful insemination of estrous females. The effects of sperm plug displacement on paternity in Lemur catta are unknown, as no study had heretofore documented copulatory plug displacement in this species. The first-male mating advantage suggested for Lemur catta should be re-evaluated where mating order is known, and copulatory plug displacement during mating, or lack thereof, is identified. Because there is a tendency for first-mating males to mate-guard for longer periods of time in Lemur catta, the latency period between the first mate's ejaculation and that of subsequent mates may be an important determinant of male fertilization success.     

Mating status regulates post‐mating refractory period in Striacosta albicosta females

In polyandrous insect species, males may transfer substances to reduce sperm competition by affecting female sexual receptivity. In this study, we determined the incidence of polyandry in females of Western bean cutworm (WBC), Striacosta albicosta (Smith) (Lepidoptera: Noctuidae), and investigated the influence of both previous female and male mating history on the duration of mating, the female refractory period, and subsequent calling behavior of females under controlled laboratory conditions. The mating status of WBC males influenced mating duration, with copulations involving previously mated males taking longer, possibly related to the time required to produce an ejaculate. The duration of the female refractory period and the onset time of recalling during the scotophase were both affected by female mating history, but not by that of the males. Females had a shorter refractory period and resumed calling activity earlier after their second and third matings than after their first mating. The earlier onset of calling by previously mated females could reduce competition with virgin females and their shorter refractory period could explain the high incidence of polyandry observed in nature.     

Sex Differences in Mate Preference Between Two Hybridizing Species of Poeciliid Fish

When hybridization between two species in secondary contact is costly, natural selection should favor pre‐mating isolation barriers. The invasive Gambusia affinis has been introduced to habitats of a closely related species, the endangered Gambusia nobilis. Although other Gambusia species readily hybridize in secondary contact, previous studies in this system found low abundance of hybrids in sympatry. To examine whether hybridization is limited by behavioral pre‐mating isolation that may have evolved in allopatry, I examined each species’ mating preferences using individuals from allopatric populations in male and female visual/olfactory association preference tests as well as open mating tests with and without male–male competition. Gambusia affinis and G. nobilis males had significant association preference for conspecific females in visual/olfactory tests. Only G. nobilis females had statistically significant preference for conspecific males. In open mating tests, males of both species had lower chase times overall when in competition, but there was no difference in number of copulation attempts. Males of both species had higher copulatory success rates with conspecific females when in competition, suggesting females may exert some control over copulation success of males. These results suggest that there are differences in mating preferences between these species. This mate choice may act as a pre‐mating isolating barrier to reduce hybridization in sympatry, a proposed threat to the endangered G. nobilis.     

Costs of Mate Guarding and Opportunistic Mating Among Wild Male Japanese Macaques

A trade-off relationship between mating and feeding effort is important when considering reproductive strategies of long-lived species. I compared the influence of male sexual activities, female mate-choice behaviors and the daily activity budget on male mating success among males in a group of wild Japanese macaques (Macaca fuscata yakui) on Yakushima Island. The 1st-ranking male, which had immigrated into the troop at this rank, more frequently approached peri-ovulatory females, spent more time grooming peri-ovulatory females and in mounting series and spent less time feeding than subordinate males did. The 1st-ranking male attained the highest mating success as a result of his high expenditure of time and energy in sexual behaviors directed toward peri-ovulatory females. Mating success of subordinate males did not relate to the amount of sexual effort, but instead to the frequency of female approaches, female rush toward males and the number of peri-ovulatory females within the group. The pattern of intermale competition shifted from nearly contest competition to scramble competition as the number of peri-ovulatory females in the group increased. Feeding time of subordinate males did not vary between the days when they copulated and the days when they did not. The findings demonstrate that mate guarding in the 1st-ranking male is a high-cost mating tactic, while opportunistic mating in subordinate males is a low-cost mating tactic. The differences in male mating tactics are probably related to male life history and to the formation of groups with a high socionomic sex ratio.     

Timing of female sexual unreceptivity and male adjustment of copulatory behaviour under competition risk in the wolf spider <Emphasis Type="Italic">Schizocosa malitiosa</Emphasis>

Males can change their copulatory or sperm transfer patterns in response to sperm competition risk. Schizocosa malitiosa performs long copulations, which include two consecutive patterns (Patterns 1 and 2). Virgin females are very sexually receptive, but mated females diminish their receptiveness. Female unreceptivity has been attributed to the action of receptivity-inhibiting substances, mainly transferred during Pattern 1. We tested: (1) if females who mated only with Pattern 1 were immediately unreceptive; (2) male and female behaviours when the copulating couple was exposed to another male. For (1), we interrupted mating when Pattern 1 finished and immediately exposed the female to a second male. For (2), we introduced a second male when the couple was starting (Ei) or finishing copulation (Li). Females were unreceptive immediately after finishing Pattern 1. Males from Ei and Li dismounted and approached the second males. Ei males diminished the frequencies of insertion after perceiving the presence of a second male and dismounted less frequently when copulating with heavy females. The study provides insights about the timing of sexual unreceptivity in S. malitiosa under possibilities of sperm competition, discussing male adjustment of copulatory behaviour in the presence of rival males. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Ecdysteroids as a component of the genital sex pheromone in two species of hard ticks,Dermacentor variabilis andDermacentor andersoni (Acari: Ixodidae)

Neutering of part-fed females eliminated copulatory behavior inDermacentor variabilis (Say) andD. andersoni Stiles males. Extracts from the anterior reproductive tracts of part-fed (7 days) females partly restored the male copulatory behavior in conspecific neutered females. Similar extracts from unfed females did not restore the behavior, suggesting that the pheromone was produced during feeding. Perception of the genital sex pheromone by sensillae on the male cheliceral digits was confirmed by electrophysiological techniques.Males ofD. variabilis andD. andersoni responsed positively to authentic ecdysone and 20-hydroxyccdysone (20HE) in neutered female bioassays. Responses to sterols were significantly lower than to ecdysteroids. Electrophysiological assays suggest a sensitivity of males to high doses of ecdysteroids. The strongest responses were to 20HE in both species. Ecdysteroids, specifically ecdysone and 20HE were shown to be present in the anterior reproductive tracts in excess of amounts that could be explained by mere hemolymph contamination. Ecdysteroids were also found in washings of the vaginal lumen of these two species.Dermacentor andersoni females contained larger amounts of ecdysteroids thanD. variabilis females. 20-hydroxyecdysone and possibly ecdysone appear to be components of the genital sex pheromone ofD. variabilis andD. andersoni. Species recognition may be facilitated by these components, but the complete mechanism is not yet fully understood.Supported by grant AI10, 986 from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, U.S. Public Health Service, Department of Health and Human Services, Bethesda, Maryland.