Parental behaviour in relation to the occurrence of sneaking in the common goby

Parental behaviour in relation to the occurrence of sneaking in the common goby

ANIMAL BEHAVIOUR, 1998, 56, 175–179 Article No. ar980769 Parental behaviour in relation to the occurrence of sneaking in the common goby OLA SVENSSON...

87KB Sizes 114 Downloads 43 Views

ANIMAL BEHAVIOUR, 1998, 56, 175–179 Article No. ar980769

Parental behaviour in relation to the occurrence of sneaking in the common goby OLA SVENSSON*, CARIN MAGNHAGEN†, ELISABET FORSGREN* & CHARLOTTA KVARNEMO*

*Department of Zoology, Uppsala University †Department of Aquaculture, Swedish University of Agricultural Sciences (Received 20 January 1997; initial acceptance 3 March 1997; final acceptance 17 November 1997; MS. number: 5456R)

ABSTRACT To investigate whether male common gobies, Pomatoschistus microps (Pisces, Gobiidae), treat their offspring differently depending on confidence of paternity, we conducted an experiment in which randomly chosen males either spawned alone with a female, or with a sneaking male present. Males did not treat their brood differently whether they had experienced sneaking or not. Our estimates of parental care, nest defence against potential egg predators and fanning rate were the same for the two treatments. Furthermore, there was no difference in filial cannibalism (eating their own progeny) between males that had been sneaked upon and males that had not. However, nest-guarding males that ate some of their brood had a smaller original brood area than other males. This suggests either an increase in paternal expenditure with increased brood size or a threshold value (absolute brood size or proportion of nest space covered) above which males do not cannibalize eggs. 

paternity (reviewed in Owens 1993): (1) availability of information on paternity; (2) variation in the risk of cuckoldry between breeding attempts; and (3) trade-offs concerning costs and benefits with increasing or decreasing parental care. To apportion care to the brood in proportion to his share of paternity, the male must be able to assess this proportion, for example by discriminating between related and nonrelated offspring. This seems feasible, as female guppies, Poecilia reticulata, preferentially prey on unrelated fry (Loekle et al. 1982) and female three-spined sticklebacks, Gasterosteus aculeatus, prefer to attack unrelated eggs (FitzGerald & van Havre 1987). Males of the fathead minnow, Pimephales promelas, appear to discriminate between adopted and their own eggs (Sargent 1989) with adopted eggs suffering higher mortality because of less care and defence. In a comparison of data for 52 species of bird, Møller & Birkhead (1993) found that the extent of paternal care during the feeding period was negatively related to the extrapair paternity, in a cross-taxonomic correlation controlled for phylogenetic associations. In the reed bunting, Emberiza schoeniclus, males give more care to broods in which they possess a higher proportion of paternity than to ones where their share is less; however, males also feed broods consisting of only extrapair young, suggesting that they cannot recognize their own offspring (Dixon et al. 1994). A male may estimate his probable paternity if it varies in a predictable rather than in a random manner, for

Alternative reproductive tactics have evolved in many fish species where reproductive success depends on aggression and competition between males (Dominey 1981, 1984). Usually smaller males, which are less likely to defend a nest and attract a female successfully, act either as sneakers (interfere with spawning pairs, fertilize some of the eggs and leave) or as satellites (tolerated by the dominant male and sometimes even help with territory defence; Taborsky et al. 1987). In the common goby, Pomatoschistus microps, male reproductive behaviour is dependent on size. Males of medium size act both as sneakers and nest builders, while the smallest males do not try to build nests even when alone and the largest males never try to sneak fertilizations (Magnhagen 1992). Parental care is often a costly part of reproduction (reviewed in Smith & Wootton 1995), and hence, natural selection should favour males that avoid caring for unrelated young (Trivers 1972). Theoretical work focusing on the relationship between paternity and male parental care suggests that there are three main factors determining whether the male will adjust his effort in response to Correspondence: C. Magnhagen, Department of Aquaculture, SLU, S-901 83 Umea˚, Sweden (email: [email protected]). O. Svensson is at the Department of Zoology, Uppsala University, Villava¨gen 9, S-752 36 Uppsala, Sweden. E. Forsgren is now at the Kristineberg Marine Biology Station, S-450 34 Fiskeba¨ckskil, Sweden. C. Kvarnemo is now at the Department of Zoology, University of Western Australia, Nedlands, WA 6009, Australia. 0003–3472/98/070175+05 $30.00/0

1998 The Association for the Study of Animal Behaviour

175



1998 The Association for the Study of Animal Behaviour

176 ANIMAL BEHAVIOUR, 56, 1

example, with season or age (Westneat & Sherman 1993). The most accurate way for a male to assess his likelihood of paternity, however, occurs during the female’s fertile period. Studies on the relation between paternity and male care of the offspring has to a great extent been focused on birds (e.g. Møller & Birkhead 1993). To our knowledge, this problem has never been studied experimentally in fish, even though paternal care, as well as sneak fertilizations, are common in many fish species (Taborsky 1994; Gross 1996). Our aim in the present study was to test experimentally if males of the common goby treat their offspring differently after having experienced sneaking compared with when no sneaking occurred. The behaviours studied were fanning and defence of the nest against egg predators. We also observed the occurrence of filial cannibalism, that is, whether the male ate some of the eggs. METHODS

Study Species The common goby is a small marine fish, living in shallow, soft bottom areas. It is short-lived, reproducing repeatedly during only one season, from May to August (Miller 1975). The male builds a nest using a mussel shell, which he covers with sand, and he fans and guards the eggs until hatching. Males may receive eggs from several females; large males are usually more successful in occupying large mussel shells, and also obtain more eggs in their nests (Magnhagen & Vestergaard 1993). The foraging ability of the males is restricted by the demands of egg guarding (Magnhagen 1986). This together with care of multiple clutches are conditions that, according to Rohwer (1978), are expected to favour the evolution of filial cannibalism. Male common gobies are known to cannibalize their own brood (Magnhagen 1992; Jones 1997). Sneak mating also occurs, with relatively small males entering other males’ nests during spawning in attempts to fertilize some of the eggs (Magnhagen 1992).

Procedure We conducted the experiments during June and July 1995 at Klubban Biological Station on the west coast of Sweden (5815 N, 1128 E). Male and female common gobies were caught in the nearby Kilviken bay with a hand trawl. The experiment was carried out indoors with continuously renewed sea water from a depth of 4 m, with a salinity of 20–25‰, and water temperatures ranging between 16 and 18.5C during paternal care. Water temperature, salinity and light followed natural conditions. There was a 2-cm layer of fine sand in all aquaria. Clay flowerpots cut in half (width 4.5 cm, depth 4 cm) were used as standardized nest sites. All males and females in the observation aquaria were fed with chopped mussel, Mytilus edulis, meat in excess every third day. Uneaten mussel meat was removed the day after feeding. Fish kept in storage tanks were fed with mussels every

day. After the experiment, we released all the fish into their natural habitat. To investigate whether males adjust their parental effort depending on their confidence of paternity, we divided 25 males into two groups of 10 and 15, respectively, with the same length variation in the two groups (XSE=41.00.84 cm, N=10; 40.60.73 cm, N=15). We placed each male in an 18-litre aquarium with a half flowerpot 20 cm from the front. The aquaria were screened off to prevent males from seeing each other. Each of the males in the group of 15 males was assigned a potentially sneaking male (30–36 mm total body length and at least 4 mm shorter than the nest-holding male) which was added to the aquarium after a nest was built. We placed a female in a net cage (16.512.5 12.0 cm) in each aquarium to stimulate the male to start building a nest and removed her as soon the nest was ready. Any male that had not built a nest after 7 days was exchanged with another male of the same size. When a male had built a nest, we introduced a new female into the aquarium. She was allowed to swim freely and spawn with the male in the nest. We observed the aquaria with an additional male continuously before and during spawning to see if sneaking occurred. If a female did not respond to the male’s courtship, we replaced her with a new female after 3 h. The next day, we removed the sneaking male and the female. We also removed the pots briefly, and marked out the area covered with eggs with a pencil. Then the pots were returned and all the males rebuilt their nests. To estimate parental expenditure, we measured fanning time and defence of the nest. The reproductive value of a brood increases as it develops (Sargent & Gross 1985) and common gobies are known to defend and fan their broods more at the end of a reproductive cycle (Magnhagen & Vestergaard 1993; Jones 1997). Thus, to get as much difference between treatments as possible these measurements were done the day before hatching, that is, 5 days after spawning. Fanning time was measured as total fanning time per 10 min, starting when the male began a fanning bout. We used a torch to observe the males in the nests. To measure the defence of the nest against egg predators, we used the following procedure. A dead shore crab, Carcinus maenas, with a carapace width of 13 mm, was attached to a thin iron rod 40 cm long. We moved the crab slowly towards the nest, starting from the midpoint of the front of the aquarium. The male’s immediate response to the threat was observed and scored on a five-point scale. (1) The male was far away from the nest all the time, or escaped from the nest as soon as the crab was put into the water. (2) The male escaped from the nest as the crab approached. (3) The male was in the nest all the time, or swam out and watched the crab. (4) The male attacked and/or spread his fins and gills, and/or made tail-beats. (5) The male performed several aggressive attacks, and/or bit the crab repeatedly and/or tried to move the crab away.

SVENSSON ET AL.: GOBY PARENTAL BEHAVIOUR 177

Table 1. Estimates of parental care (X±SE) in male common gobies from two treatments without and with sneaking by another male, respectively Male trait Partial brood loss area (cm2) Partial brood loss (%) Fanning time (s) Defence score

No sneaking

Sneaking

t/Z

df/N

P

0.58±0.33 11.6±3.7 309±33 7.5±0.64

0.34±0.23 9.9±3.3 270±57 7.8±0.46

0.56 0.26 0.61 0.29

17 17 15 10:9

0.58 0.80 0.51 0.77

The differences were tested with a t test. Percentages were arcsine square root-transformed prior to testing. The defence score was tested with a Mann–Whitney test (N=no sneaking:sneaking).

RESULTS Sneaking was observed in nine of the 15 cases where the male had been placed with a potential sneaker male. All males trying to sneak fertilizations successfully entered the nest. The sneaker was seen to enter one to three times and stayed in the nest between 28 s and more than 1 h (median 88.5 s). In the other six aquaria where a potential sneaker was present, no sneaking behaviour was ever seen, as five potential sneakers buried themselves in the substrate (a natural behaviour for inactive common gobies) during the matings and the sixth tried to attract the female to one corner of the aquarium. There was no difference in female size or egg area in the replicates with no sneaker present, other male present but no sneaking observed, and sneaking observed, respectively (female length: F2,13 =0.65, P=0.54; egg area: F2,22 =1.95, P=0.17). Therefore, we excluded these six cases from the data set when looking at the effect of exposure to sneaking.

3 Egg area eaten (cm2)

Immediately afterwards, we carried out the same procedure using a live netted dogwhelk, Hinia reticulata, 15 mm long, whose shell was attached with superglue to a 40-cm iron rod. The two scores were added to yield a total defence score. Directly after these observations, we removed the pot, and if eggs had disappeared since spawning, marked the reduced brood area with a pencil. Common goby males are known sometimes to cannibalize some of their own brood (Magnhagen 1992; Jones 1997), and the egg losses in this study were most likely due to filial cannibalism. If the losses were evenly dispersed over the egg mass, we estimated the total loss as a percentage of the original brood. The egg areas, serving as estimates of the number of eggs in each brood, were traced on to transparent paper, cut out and weighed. By calibration with the weight of two pieces of paper, measuring 100100 mm, we could calculate the area of the brood on days 1 and 6. In the statistical analyses, we used three measurements of egg losses: (1) whether there were any egg losses; (2) the proportion of original brood area lost (as a percentage); and (3) the absolute area of brood loss. Proportions were arcsine (square root) -transformed prior to analyses. For looking at differences in defence score we used a nonparametric test since this is a noncontinuous scale. All test probabilities are two-tailed.

2

1

0

5

6

7

8

9

10

11

12

13

14

15

16

Initial egg area (cm2) Figure 1. Brood area lost to filial cannibalism in relation to initial brood area in nests of the common goby. Note that some points are overlapping.

The occurrence of sneaking did not affect filial cannibalism, as three of the nine males that experienced sneaking, and three of the 10 males that did not, lost some of their eggs (Fisher’s exact test: P=1.00). There was no difference between the two treatments in absolute or proportional loss of eggs (Table 1). Furthermore, neither the fanning time nor the defence score differed between treatments (Table 1). However, the seven males that lost some of their brood (including one male that was not considered in the sneak–nonsneak comparison) had a significantly smaller original brood area (XSE= 8.50.6 cm2) than the 18 males that did not lose some of their brood (10.40.5 cm2; t23 =2.17, P<0.05; Fig. 1). No male with a brood area over 11 cm2 lost any eggs. No male lost his whole clutch. There were no significant correlations between egg area and defence score (rS =0.02, N=22, P=0.93) or fanning rate (rS = 0.14, N=24, P=0.90), nor between male size and defence score (rS =0.14, N=22, P=0.52) or fanning rate (rS =0.34, N=24, P=0.10). When looking only at the males that had experienced sneaking, there was no correlation between the sneaker’s time in the nest and defence score (rS = 0.27, N=9, P=0.49), fanning rate (rS =0.28, N=9, P=0.49) or egg area lost (rS =0.02, N=9, P=0.95). DISCUSSION Our results show that male common gobies did not treat their brood differently whether or not they had

178 ANIMAL BEHAVIOUR, 56, 1

experienced sneaking. There were no differences in partial brood losses between the treatments. Furthermore, there were no differences in fanning time or defence of the nest late in the parental cycle when one can expect the highest investment in fanning time and rate of defence (Sargent & Gross 1985). Thus, we found no support for the hypothesis that males reduce parental care in relation to a presumed lower confidence of paternity. A male should take a higher risk with increasing value of the brood. In other studies of the common goby, risk taking and defence increased with developmental stage of the eggs, and defence behaviour was also dependent on brood size (Magnhagen & Vestergaard 1991, 1993). The fact that we found no difference between treatments, with and without sneaking, in the males’ defence of the nest against potential egg predators suggests that the males gave their brood the same value regardless of treatment. Defence of the eggs should reflect not only a male’s willingness to take a risk but also how much energy he is willing to allocate to their defence. This was apparently not affected by the occurrence of sneak maters nor by how long the sneaker spent in the nest. Similarly, some recent studies have shown parental care among birds does not vary in relation to paternity (Whittingham et al. 1993; Whittingham & Lifjeld 1995). Our lack of correlation between the occurrence of sneaking and parental effort could be because (1) the males cannot assess paternity, or (2) the benefits of decreasing brood care or of cannibalizing the eggs may be lower than the cost in terms of mortality of the male’s own offspring. Reynolds (1996) argued that in substrate-spawning fish, especially those with care provided by the male alone, one might expect that low cost per brood would lead to reduced correlations between certainty of paternity and care. Furthermore, if the male’s relatedness to the offspring, under natural conditions, is constant or varies in an unpredictable way, there would be no selection for a behavioural response to sneaking, assuming that the male is incapable of identifying eggs fertilized by the sneaking male. There is also a possibility of a threshold response to paternity. Whittingham et al. (1993; but see Houston 1995) predicted a continuous (gradual) decline in male parental care in monogamous birds in response to decreased paternity when the relationship between male parental care and offspring recruitment is concave-down, but a threshold relationship between male parental care and paternity when the relationship is S-shaped. The males should then lower or terminate their care only if their paternity confidence is low. If the sneaking males in our study fertilized only a small proportion of the eggs in the brood, the paternity level of the nest-guarding male may still have been above the critical threshold. Thus, one reason for failing to find any treatment effect could be that sneaking males were not very successful in fertilizing the eggs. In the bluegill sunfish, Lepomis macrochirus, protein electrophoresis showed that up to 59% of a brood could be sired by cuckolder males (Philipp & Gross 1994), while, according to DNA fingerprinting, on average 13.5% of stickleback fry had a father other than the guarding male (Rico et al. 1992). In gobiid

species, the males turn upside down and release sperm trails, flat bands of viscous material in which sperm are embedded, on the nest surface before the female starts laying eggs (Marconato et al. 1996; Ota et al. 1996). Common goby males also turn upside down early in the courtship sequence (personal observation), and may very likely release sperm which may affect the fertilization success of sneaking males. However, since the second male entered several times or stayed in the nest for up to 1 h it is very likely that he managed to fertilize some of the eggs. To measure the fertilizing success of the sneaking males, DNA fingerprinting could be used. The observation that the nest-guarding males that ate some of their clutch had a smaller original brood area, always less than 11 cm2, suggests either an increase in paternal expenditure with increased brood size or that males do not cannibalize broods over a certain area, or proportion of the largest potential brood of the nest. The latter can be explained by males caring for multiple clutches: males whose nests are not full still try to attract females (personal observation) and are therefore able to compensate for the egg losses, while males with a full nest may not try to attract females. The size of the brood thus seems to be a better estimate of brood value than the risk of reduced paternity in the presence of a sneaker. A male may actually consume his entire brood if it is exceptionally small (O. Svensson, E. Forsgren & C. Kvarnemo, unpublished data). Similar results have also been found in the sand goby, Pomatoschistus minutus (Forsgren et al. 1996), and in the fantail darter, Etheostoma flabellare, where males with small egg masses ate all of their eggs whereas males with large egg masses were only partial cannibals (Lindstro ¨ m & Sargent 1997). The lack of full clutch cannibalism in our study is, most probably, explained by the lack of very small clutches. However, a male with no need of an additional food source may not cannibalize his clutch at all. In a study on the common goby, males on a low food ration ate parts of their clutches much more often than males on a high food ration (O. Svensson, E. Forsgren & C. Kvarnemo, unpublished data). This may explain why the brood areas of the males where no cannibalism occurred ranged from almost the smallest to the largest, as their body condition might have been better when they were caught, compared with the cannibalistic males. To conclude, male common gobies did not treat their brood differently whether they had experienced sneaking or not. Nest defence and fanning rate were the same, and there was no difference in filial cannibalism between males that had been subject to sneaking and males that had not. Instead, the fact that males that cannibalized some of their brood had a smaller original brood area than males that did not suggests that brood size serves as a cue for the male’s parental expenditure. Acknowledgments We thank I. Ahnesjo ¨ , S. Ulfstrand and the anonymous referees for valuable comments on the manuscript. Financial support was provided by Uppsala University (to O.S.) and Lennander’s foundation (to C.K. and E.F.).

SVENSSON ET AL.: GOBY PARENTAL BEHAVIOUR 179

References Dixon, A., Ross, D., O’Malley, S. L. C. & Burke, T. 1994. Paternal investment inversely related to degree of extra-pair paternity in the reed bunting. Nature, 371, 698–700. Dominey, W. J. 1981. Maintenance of female mimicry as a reproductive strategy in bluegill sunfish (Lepomis macrochirus). Environmental Biology of Fishes, 6, 59–64. Dominey, W. J. 1984. Alternative mating tactics and evolutionarily stable strategies. American Zoologist, 24, 385–396. FitzGerald, G. J. & van Havre, N. 1987. The adaptive significance of cannibalism in sticklebacks. Behavioral Ecology and Sociobiology, 20, 125–128. Forsgren, E., Karlsson, A. & Kvarnemo, C. 1996. Female sand gobies gain direct benefits by choosing males with eggs in their nests. Behavioral Ecology and Sociobiology, 39, 91–96. Gross, M. R. 1996. Alternative reproductive strategies and tactics: diversity within sexes. Trends in Ecology and Evolution, 11, 92–98. Houston, A. I. 1995. Parental effort and paternity. Animal Behaviour, 50, 1635–1644. Jones, J. C. 1997. Trade-offs in fish reproduction: responses of the common goby to oxygen stress. Ph.D. thesis, University of East Anglia. Lindstro ¨ m, K. & Sargent, R. C. 1997. Food access, brood size and filial cannibalism in the fantail darter, Etheostoma flabellare. Behavioral Ecology and Sociobiology, 40, 107–110. Loekle, D. M., Madison, D. M. & Christian, J. J. 1982. Time dependency and kin recognition of cannibalistic behaviour among poeciliid fishes. Behavioral and Neural Biology, 35, 315–318. Magnhagen, C. 1986. Activity differences influencing food selection in the marine fish Pomatoschistus microps. Canadian Journal of Fisheries and Aquatic Sciences, 43, 223–227. Magnhagen, C. 1992. Alternative reproductive behaviour in the common goby, Pomatoschistus microps: an ontogenetic gradient? Animal Behaviour, 44, 182–184. Magnhagen, C. & Vestergaard, K. 1991. Risk taking in relation to reproductive investment and future reproductive opportunities; field experiments on nest-guarding common gobies, Pomatoschistus microps. Behavioral Ecology, 2, 351–359. Magnhagen, C. & Vestergaard, K. 1993. Brood size and offspring age affecting risk-taking and aggression in nest-guarding common gobies. Behaviour, 125, 233–243. Marconato, A., Rasotto, M. B. & Mazzoldi, C. 1996. On the mechanism of sperm release in three gobiid fishes (Teleostei, Gobiidae). Environmental Biology of Fishes, 46, 321–327.

Miller, P. J. 1975. Age-structure and life-span in the common goby. Journal of Zoology, 177, 425–448. Møller, A. P. & Birkhead, T. R. 1993. Certainty of paternity covaries with paternal care in birds. Behavioral Ecology and Sociobiology, 33, 261–268. Ota, D., Marchesan, M. & Ferrero, E. A. 1996. Sperm release behaviour and fertilization in the grass goby. Journal of Fish Biology, 49, 246–256. Owens, I. P. F. 1993. When kids just aren’t worth it: cuckoldry and parental care. Trends in Ecology and Evolution, 8, 269–271. Philipp, D. P. & Gross, M. R. 1994. Genetic evidence for cuckoldry in bluegill Lepomis macrochirus. Molecular Ecology, 3, 563–569. Reynolds, J. D. 1996. Animal breeding systems. Trends in Ecology and Evolution, 11, 68–72. Rico, C., Kuhnlein, U. & FitzGerald, G. J. 1992. Male reproductive tactics in the threespine stickleback: an evaluation by DNA fingerprinting. Molecular Ecology, 1, 79–87. Rohwer, S. 1978. Parent cannibalism of offspring and egg raiding as a courtship strategy. American Naturalist, 112, 429–440. Sargent, R. C. 1989. Allopaternal care in the fathead minnow, Pimephales promelas: stepfathers discriminate against their adopted eggs. Behavioral Ecology and Sociobiology, 25, 379–385. Sargent, R. C. & Gross, M. R. 1985. Parental investment decision rules and the Concorde fallacy. Behavioral Ecology and Sociobiology, 17, 43–45. Smith, C. & Wootton, R. J. 1995. The cost of parental care in teleost fishes. Review of Fish Biology and Fisheries, 5, 7–22. Taborsky, M. 1994. Sneakers, satellites, and helpers: parasitic and cooperative behavior in fish reproduction. Advances in the Study of Behavior, 23, 1–100. Taborsky, M. I., Hudde, B. & Wirtz, P. 1987. Reproductive behaviour and ecology of Symphodus (Crenilabrus) ocellatus, a European wrasse with four types of male behaviour. Behaviour, 102, 82–118. Trivers, R. L. 1972. Parental investment and sexual selection. In: Sexual Selection and the Descent of Man 1871–1971 (Ed. by B. Campbell), pp. 136–179. London: Heinemann. Westneat, D. F. & Sherman, P. W. 1993. Parentage and the evolution of parental behaviour. Behavioral Ecology, 4, 66–77. Whittingham, L. A. & Lifjeld, J. T. 1995. High paternal investment in unrelated young: extra-pair paternity and male parental care in house martins. Behavioral Ecology and Sociobiology, 37, 103–108. Whittingham, L. A., Dunn, P. O. & Robertson, R. J. 1993. Confidence of paternity and male paternity and male parental care: an experimental study in tree swallows. Animal Behaviour, 46, 139–147.