Behavioral Ecology Vol. 14 No. 3: 374-381
© 2003 International Society for Behavioral Ecology
Effects of a low oxygen environment on parental effort and filial cannibalism in the male sand goby, Pomatoschistus minutus
Department of Zoology, Stockholm University, SE-106 91 Stockholm, Sweden
Address correspondence to M. Liss
ker. E-mail: maria.lissaker{at}zoologi.su.se.
Received 29 January 2002; revised 28 August 2002; accepted 28 August 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
In fish, brood cycling parental males sometimes eat some or all of their eggs, a behavior termed filial cannibalism. We tested predictions of filial cannibalism models related to the cost of parental care in the male sand goby, Pomatoschistus minutus, by increasing the parental effort (fanning expenditure) through reduced levels of dissolved oxygen to 39% in an experimental group, whereas a control group had fully saturated water. Males showed both full-clutch cannibalism and partial-clutch cannibalism in both treatments. Giving the males one to three females to spawn with, we found that small clutches were completely eaten more often than were larger ones, whereas partial-clutch cannibalism was not affected by clutch size. Although treatment did not affect filial cannibalism, it did affect a male's energy state such that males in the low oxygen treatment lost more body fat, indicating a greater fanning effort. This shows that males in the low oxygen treatment allocated more energy to the present brood, potentially at the expense of future reproductive success. Our study strongly suggests that filial cannibalism in male sand gobies represents a strategic life-history decision as an investment in future reproductive success, and is not triggered by a proximate need for food necessary for the male's own survival. Furthermore, males in the low oxygen treatment built nests with larger entrances, and were less likely to rebuild their nests after destruction. Presumably, this makes fanning easier but the nest more vulnerable to predators, suggesting a trade-off between fanning and nest defense.
Key words: clutch size, dissolved oxygen, fanning, filial cannibalism, nest building, parental effort, Pomatoschistus minutus, reproductive success, sand goby.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The cost of reproduction, which includes cost of survival and cost of future reproduction, represents one of the most prominent life-history trade-offs (Stearns, 1989
). As different animals have different physiological and life-history constraints, a diversity of parental care patterns and mating systems have evolved. Among teleost fishes, most families have no parental care, and in those families in which parental care is provided, this is usually performed by the male only (Gross and Shine, 1981). Parental care often involves fending off predators and fanning, as well as keeping the offspring healthy by removing dead or deceased eggs (Clutton-Brock, 1991). All these activities are energetically costly, and in some species of fish, this has promoted the evolution of filial cannibalism.
Filial cannibalism is the commonly used expression for when guarding males consume some or all of their own genetic offspring (Rohwer, 1978). This is a complex subject involving individual conflicts of interest between costs (parental investment) and benefits (reproductive success), as well as conflicts between the sexes (Rohwer, 1978). It can also be viewed as a highly adaptive mating strategy, allowing individuals to prolong their breeding season and invest in future reproduction by sacrificing some or all of the present brood (Sargent, 1992). Filial cannibalism is known to occur in many paternally caring fish species, and has been studied in, e.g., fantailed darter, Etheostoma flabellare (Lindström and Sargent, 1997); three-spined stickleback, Gasterosteus aculeatus (Belles-Isles and FitzGerald, 1991); Cortez damselfish, Stegastes rectifraenum (Hoelzer, 1992); the blenny Aidablennius sphynx (Kraak and Weissing, 1996); sand goby, Pomatoschistus minutus (Forsgren et al., 1996; Lindström, 1998); common goby, Pomatoschistus microps (Kvarnemo et al. l998); and river bullhead, Cottus gobio (Marconato et al., 1993).
Rohwer (1978) proposed the first energy-based model of filial cannibalism, in which the general conditions favoring the evolution of this strategy are (1) that the male can guard multiple clutches, (2) that the demands of egg guarding restrict the male's foraging, and, obviously, (3) that he benefits more than he loses from egg eating. Long incubation periods will intensify the fasting period and, hence, the potential benefit from egg eating in terms of reduced risk of starvation. Another factor that may be of influence is body size and condition of the male. A small body size will raise the cost-to-benefit ratio because the amount of fat in small fish is less likely to cover the energy cost of more than one brood cycle (Rohwer, 1978).
Sargent (1992) developed this model further and made predictions about under which circumstances filial cannibalism would be adaptive. A large clutch has a higher reproductive value than does a small clutch because the energetic cost of continued care for a small clutch may not be outweighed by the benefits. Hence, smaller clutches should be completely eaten more often than are larger ones. On the other hand, the cost of caring could increase with brood size in species that actively oxygenate their eggs, because this form of care is not shared (Perrin, 1995). Further, it is logical to assume that full-clutch filial cannibalism represents only an investment in future reproduction, whereas partial filial cannibalism can influence both present and future reproduction (Sargent, 1992). In this case, Sargent (1992) proposed that filial cannibalism represents a decision by the parent. It has been suggested, by Marconato et al. (1993), that the male river bullhead is forced to eat the eggs simply in order to stay alive while brooding. However, as Manica (2002) points out, eating eggs could also lead to better survival chances during the winter and, hence, could be seen as an investment in future reproductive success. Manica (2002) also argues that in partial filial cannibalism, the number of eggs eaten by the parent should increase with brood size, as the cost of eating one egg decreases proportionally with an increase in brood size. This should hold until a threshold level, at which the male is satiated and the value of eating is diminished (Manica, 2002). Another explanation for filial cannibalism is that removal of sick and deceased eggs may incite the male to consume healthy eggs (Sargent, 1992). It has also been suggested that eggs are rich in certain nutrients important to the male (Belles-Isles and FitzGerald, 1991; Sargent, 1992). However, Kvarnemo et al. (1998) tested this hypothesis in common gobies, with one group fed eggs plus food and one group fed only food. Both groups showed filial cannibalism to the same extent, which indicates that at least in the common goby, filial cannibalism occurs not because of a lack of any important nutrients in their diet.
Further, in three-spined sticklebacks, which raid the nests of conspecifics and eat the eggs, females may join in raids toward nests in which their own eggs are laid, but they never initiate the raid (FitzGerald and van Havre, 1987). This could represent the best alternative in a bad situation, considering that the eggs would be eaten anyway and the female is allowed to regain some of the energy invested in producing the eggs. Finally, the aggressively territorial male pupfish, Cyprinodon pecosensis, will eat his own eggs if he detects them, but if they escape notice from the male, they are well protected within his territory (Kodric-Brown, 1983).
In shallow marine and estuarine habitats, periods of low levels of dissolved oxygen are not uncommon (Baden et al., 1990; Breitburg, 1992; Jones and Reynolds, 1999a). This could be caused either by natural factors, such as tidal influences and nocturnal algae respiration, or by high biochemical oxygen demand as a result of eutrophication or even high organic input by industries. In our study area, eutrophication promotes excessive growth of filamentous green algae in shallow coastal areas during late spring and early summers (Kautsky et al., 1986; Pihl et al., 1996, 1999). The breakdown of such large amounts of organic material may then lead to hypoxic conditions in the areas affected(Baden et al., 1990).Low levels of dissolved oxygen can affect the availability of energy for locomotion, growth, and reproduction in fishes and, in severe cases, lead to death (Kramer, 1987; Petersen and Petersen, 1990). It is clear that oxygen supply by a fanning parent is of substantial importance for development and survival of the offspring (Fonds and van Buurt, 1974; Kramer, 1987; Takegaki and Nakazono, 1999; Zoran and Ward, 1983). Several studies related to the subject have shown an increase in both male fanning frequency and in the time a male spends fanning the eggs as dissolved oxygen decreases (Jones and Reynolds, 1999a; Takegaki and Nakazono, 1999; Torricelli et al., 1985). Thus, lowered oxygen availability clearly increases parental expenditure, which would be likely to translate as an increased parental cost. Yet, the effects of low oxygen availability with regards to parental behavior and filial cannibalism in fishes have, so far, not been investigated thoroughly.
In this study we manipulated levels of dissolved oxygen and clutch sizes to test how the cost of fanning affects filial cannibalism in male sand gobies. With reference to the models of Rohwer (1978) and Sargent (1992), we predicted that small clutches would be completely eaten more often in a low oxygen treatment than in an environment of fully saturated water, and that more clutches would be partially eaten a in low oxygen treatment than in a control, in response to an increased parental expenditure. It also seems reasonable to assume that the threshold level at which the benefit is higher than the cost would be affected by the treatment, so that the minimum size of a clutch worth caring for would have to be larger in low oxygen in order to compensate for the higher energy expenditure of the male. Furthermore, enlargement of the nest entrance could be a way to, at least partially, compensate for the increased need of ventilation (Jones and Reynolds, 1999c, Kvarnemo, 1995). Also, physical labor of any kind is much harder under conditions of low oxygen. Therefore, we predicted that males in the low oxygen treatment ought to have larger nest entrances, and that these males would rebuild their nests to a lesser extent, after treatment started, compared with males in the control treatment.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experimental site and set-up
The sand goby, Pomatoschistus minutus, is a coastal marine fish with a life cycle of 1–2 years (Miller, 1986
). It has only one reproductive season, during which it undertakes several brooding cycles. The male builds a nest underneath an empty mussel shell and covers it with sand, leaving only a small entrance for ventilation (Guitel, 1892). He then tries to attract several females sequentially into the nest, where they attach the eggs in a single layer to the ceiling of the shell. After spawning, the female leaves, and the male provides all parental care. He defends the nest from predators, removes sick and deceased eggs, and fans the eggs to increase water circulation and, thereby, constantly provide the eggs with oxygen (Forsgren et al., 1996). This is necessary because sheltered nests to some extent reduce water circulation (Jones and Reynolds, 1999c).
We conducted this study at Tjärnö Marine Biological Laboratory on the Swedish west coast during May and June 2001. We caught sand gobies from a nearby bay by using a hand trawl. We separated them by sex and kept them in the laboratory in separate storage aquaria (approximately 130 l) with continuously flowing natural seawater (35
salinity). During this period we fed them finely cut fresh meat from mussels, Mytilus edulis, and chopped frozen common shrimps, Crangon crangon, once a day. Each catch was kept in storage aquaria for at least 3–4 days before trials.
We used 20-l aquaria covered with glass lids for the trials, each with a 3-cm layer of fine sand on the bottom. We put a halved clay flowerpot, 6 cm in diameter, in each aquarium to serve as a standardized nest site. All the aquaria were screened off from each other. The aquaria were placed on 4-cm-high boards in two plastic sinks, with 10 aquaria in each sink. The purpose was to partially submerge the aquaria in natural seawater, which was kept flowing around and underneath them to keep the temperature from rising during the treatment period.
Oxygen treatments and clutch size experiments
We randomly selected each experimental aquarium to house a high oxygen treatment (control with fully saturated sea water) or a low oxygen treatment (30–40% dissolved oxygen). Initially, however, all aquaria were provided with a continuous flow of seawater. Males with obvious mating coloration were picked out and weighed to the nearest 0.01 g and then put in the experimental aquaria, after being randomly assigned to an aquarium and, hence, a treatment. We replaced males that failed to build a nest in 3 days with new ones. The males were randomly assigned one, two, or three females with which to spawn, in order to produce a wide range of different clutch sizes. We selected the females, mature enough to spawn, and then measured them, to the nearest millimeter, for total body length before they were randomly placed with a male. Females that failed to spawn within 3 days were replaced and returned to the sea near where they were caught. The mean length of all the females used in the study was 
cm (SD).
When the females had spawned, we removed them and returned them to the sea. The clay flowerpot was removed briefly from the aquarium, and the outline of the clutch was marked with a pencil. It was then returned to the aquarium, and the male was left to rebuild the nest. After the clay flowerpots had been put back in place, we cut the water flow and the treatment started. In the aquaria of the high oxygen treatment, air was continuously bubbled into the water using air pumps, whereas in the low oxygen treatment, the dissolved oxygen level was gradually lowered to 35% over a period of 2 h by bubbling nitrogen gas into the water. The lids of the low treatment aquaria were then covered in cling wrap to prevent dissolved oxygen fluctuations. We checked the temperature and oxygen levels once daily in the high treatment aquaria and twice daily in the low treatment aquaria by using a dissolved oxygen meter (YSI 58-2). The dissolved oxygen level in the low treatment was maintained by bubbling in additional nitrogen gas when needed. Because dissolved oxygen levels depend on temperature, units of the percentage of oxygen saturation were used, rather than milligrams per liter, to ensure a consistent difference in oxygen content between the two treatments over the experimental period.
Mean temperature over the study period in the experimental aquaria was 
°C (SD) (range, 8.2°C–13.1°C) and did not differ between treatments (t test: 
, 
, 
). The mean dissolved oxygen level in the high treatment was 
% (SD) (range, 96.6–98.6%), whereas the mean dissolved oxygen level in the low treatment was 
% (SD) (range, 33.8–43.5%). Both treatments lasted 5 days. No food was given during trials. A total of 27 replicates were completed, 12 in the low oxygen treatment and 15 in the high oxygen treatment.
At the end of the treatments, we removed the clay flowerpots again and marked any area of egg loss, indicating filial cannibalism, with a pencil. Areas of initial clutch size and of filial cannibalism both were transferred to transparent film and then photocopied onto ordinary paper. These areas were then cut out and weighed to the nearest 0.001 mg. The mass of the paper gives the egg mass area, and because the eggs are attached in a single layer, the area of an egg mass is a very good estimate of the number of eggs in it (Lindström, 1998). This procedure allowed us to compare initial egg area to final egg area and determine the extent of partial and complete clutch cannibalism. When a male had completed a trial, he was removed and weighed again to the nearest 0.01 g, and the total body length was measured to the nearest mm. Across both treatments, mean initial body massof the males was 
g (SD), mean final body mass of the males was 
g (SD), and the male mean body length was 
cm (SD).
To get a more reliable indicator of parental investment than of wet body mass, the fish were frozen in labeled plastic bags at the end of the trials for later lipid extractions on each male. This was performed at the Department of Zoology at Stockholm University. We defrosted the frozen fish and dried them for 36 h at 70°C in a desiccation oven and then weighed them to the nearest 0.001 mg by using a microbalance (Cahn 28). The lipids were extracted with petroleum ether for 7 h, after which the fish were left for 2 h in a fume cupboard for the petroleum ether to evaporate. We then placed the fish in the desiccation oven again, where they were left overnight. After weighing them a second time, the lipid content of each fish was estimated by the loss of mass.
Nest building and nest entrance size
After the males had built their nests and spawned, but before treatments started, we noted the width and the height of the nest entrance. Because the nests were destroyed while we marked the size of the clutches, the males had to rebuild their nests again. This enabled us to observe which males rebuilt their nests and if there was any difference between the two treatments in nest entrance size after treatments had started.
Statistics
We tested all data for normality, skewness, and kurtosis. To achieve normality, the entire sample data used in the statistical analyses was transformed. For this purpose we used either log (x) for samples with a variance larger than the mean, square-root (x) for samples with variances that resembled a Poisson distribution, and an arsine square-root (x) transformation for observations that constituted proportions. We tested all data with the level of significance set at 
. When using ANCOVA, we tested the full model first and removed the interaction term from the model if it proved to be nonsignificant. In such cases, only the effect of the covariate(s) and factor are reported.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Twelve sand goby males were exposed to low levels of dissolved oxygen, whereas the other 15 males constituted the control group with fully saturated conditions. There was no difference in either male body length or body weight between the two treatments (Table 1). Similarly, there was no difference between treatments in male wet mass before or after trials, and there was no interaction (repeated-measures ANOVA: treatment, F1,25 = 0.51,
; time (repeated before/after), F1,25 = 0.38, 
; time x treatment, F1,25 = 0.96, 
). However, the amount of body fat after the trials was significantly lower in males in the low oxygen treatment than in males in the high oxygen treatment (Table 1). At the end of the experimental period, two clutches in the low oxygen treatment showed a delayed developmental stage of the eggs compared with two clutches reared by males in the high oxygen treatment.
|
Filial cannibalism
A total of 19 males ate some or all of their eggs, seven of the males (58%) belonged to the low oxygen treatment and 12 (80%) to the high oxygen treatment. Consequently, the number of males showing filial cannibalistic behavior did not differ between the two treatments (
2 = 1.50, N = 27, 
). Nor was there any difference in the percentage of eggs eaten between the two treatments (t test, 
, 
, 
).
Full-clutch cannibalism was shown by four (33%) males in the low oxygen treatment and by five (33%) males in the high oxygen treatment. We regarded males that had eaten more than 80% of their clutches as total-clutch cannibals because there was such a big difference in eggs eaten between these and the other partial-clutch cannibals (Figure 1). Also, because a trial period was rather short (5 days), it is likely that these males would have eaten the few remaining eggs as well before the eggs hatched. Thus, the frequency of full-clutch cannibalism was independent of treatment (2 = 0.00, N = 27, 
) and so was the size of the clutches eaten (t test, 
, 
, ) (Figure 1). Initial clutch size, on the other hand, had a significant effect on full-clutch cannibalism, such that larger clutches were less likely to be completely eaten (log-linear regression, log-likelihood = –14.36, 2 = 7.29, 
, 
) (Figure 1).
|
To test for the influence of clutch size on partial-clutch cannibalism, we used an ANCOVA on percentage of eggs eaten with oxygen treatment as the factor and initial clutch size as the covariate. Here, the nine replicates showing full-clutch cannibalism were excluded from the data set. Treatment did not affect partial-clutch cannibalism significantly (ANCOVA; treatment, F1,15 = 2.25,
), nor did initial clutch size (clutch size, F1,15 = 0.57, 
) (Figure 1). There was no significant interaction between treatment and initial clutch size. Initial clutch size was independent of treatment (t test, 
, , 
). Finally, with regards to partial-clutch cannibalism, although males in the high oxygen treatment ate more of their eggs, there was no statistical difference in percentage eggs eaten between treatments (t test, 
, 
, 
).
Energy cost
The absolute area of egg loss, estimating the amount of eggs eaten, was not correlated with a male's initial body condition, which was measured by a male's initial wet mass divided by his total body length (simple regression, 
, F1,25 = 0.85, 
). However, because treatment, body size, and the amount of eggs eaten are likely to affect a male's final body condition, we ran an analysis of covariance on male body fat (i.e., lipid content) with oxygen treatment as a factor, and male dry body mass after lipid extraction and the amount of eggs eaten as covariates. The result showed that the treatment affected the amount of body fat (ANCOVA: treatment, F1,23 = 4.74, 
). There was also a strong tendency that a male's body mass had an effect (dry weight, F1,23 = 4.07, 
) (Figure 2a), and although weaker, there was also a nonsignificant tendency that the amount of eggs eaten affected a male's amount of body fat (amount eaten, F1,23 = 3.21, 
) (Figure 2b).
|
Nest building and entrance size
Only males that built a nest within 3 days were used in the experiment. Hence, initially, males built nests to the same extent in both treatments. In contrast, after the treatments had started, only three out of 12 males in the low oxygen treatment rebuilt their nests, whereas almost all males, 13 out of 15, rebuilt their nests in the high oxygen treatment. Thus, a greater proportion of the males in the high oxygen treatment rebuilt their nests compared with the proportion of males in the low oxygen treatment (Fisher exact test,
). Of the 16 males that rebuilt their nests, all but two did so within 1 day after the treatments started. Before the treatments started, nest entrance size was not significantly different between the two treatments (t test, 
, , 
) (Figure 3). However, the final nest entrance size was significantly affected by treatment (t test, 
, , 
) (Figure 3). In addition, final nest entrance size showed no correlation with the amount of eggs eaten (Spearman rank order correlations: 
, N = 24, 
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The effects of treatment on filial cannibalism
The results in this study show that contrary to our predictions, low levels of dissolved oxygen did not affect filial cannibalism. Similar results have been the outcome of other studies in which parental costs of caring have been increased by manipulation. Lindström and Sargent (1997)
found no effect of food treatment on the fantailed darter. In a study on sand goby by Lindström (1998), a predator acted as the extra cost of caring. Again, treatment did not seem to affect filial cannibalism. On the other hand, several studies report the opposite result. Males that were given plenty of food showed less filial cannibalism than did males that received little food (Hoelzer, 1992; Kvarnemo et al., 1998). Similarly, males that had eaten food items other than their own eggs also cannibalized less than did males without other food items in their stomachs (Marconato et al., 1993). However, the negative results of our study, as well as Lindström's (1998) and Lindström and Sargent's (1997), could be explained by either too short trial periods or the fact that the males' fat reserves are at maximum levels in the beginning of reproductive season when these experiments were run (Lindström, 1998; present study). Thus, fasting might be of less importance for male physical condition in the beginning of the reproductive season than later. It could also be argued that because the males were fed before the experiments started, the males in our experiments were in a better energetic condition compared with that of males in the wild. The fact that sand goby males still show cannibalistic behavior indicates, along with Sargent's (1992) argument, that it represents a decision rather than a need. Both full-clutch cannibalism and partial-clutch cannibalism seems to represent an investment in future offspring in the case of the sand goby.
Still, it remains unclear what the benefits are from eating eggs early in the season because they would not be expected to have to cannibalize to be able to successfully complete the first brood cycle. Why doesn't the male simply complete the current brood cycle, and then forage? There are a number of reasons why it could pay off to invest in eggs received early rather than eating them. One is that early hatchlings will be larger in the beginning of the breeding season of the next year and, hence, should have a competitive advantage compared with that of later hatchlings. Another is that sand goby females prefer to lay their eggs with males that already have eggs in the nest (Forsgren et al., 1996), and thus, even a small number of eggs in the nest should be better than none. Further, it seems like a waste of time for the male to eat eggs that are already in the nest and thereby have to start over courting again, even though eating a whole small clutch seems to be a way of getting new matings sooner (Kondoh and Okuda, 2002). One explanation to why filial cannibalism occurs even early in the season seems to be associated with the age of the eggs. Rohwer (1978) predicted that males should prefer to eat the youngest eggs in the nest because they have a lower reproductive value. As a consequence, when the eggs have reached a certain stage, beyond a threshold-level, they may be "too old" in the respect that females no longer would choose to spawn next to these eggs. Consistently, females have been shown to avoid spawning in nests with old eggs (see Sikkel, 1989).
The effect of clutch size on filial cannibalism
Regarding clutch size, we found that it had a significant effect on full-clutch cannibalism but not on partial-clutch cannibalism. Only relatively small clutches were completely eaten, whereas those clutches that were partially eaten had a wider size range. According to Sargent's (1992) model, partial filial cannibalism should increase as offspring number increases and should thus represent an investment in the survival of the present brood. However, data from our present study do not support this hypothesis. Similarly, Forsgren et al. (1996) found that sand goby males with larger broods, in fact, consumed fewer eggs. The explanation for this could be that males are willing to invest more in larger, more valuable clutches, as predicted by the parental investment hypothesis (Coleman et al., 1985). Furthermore, Svensson et al. (1998) got a similar result with the common goby and suggests that this was owing either to an increase in paternal expenditure with increased brood size or to the fact that males do not cannibalize parts of 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 and are therefore able to compensate for the egg losses, whereas males with a full nest may not try to attract females. In contrast, Lindström (1998) found that sand goby males showed an increase in filial cannibalism with increasing clutch size, but this was only after he had excluded data from the large number of males who did not eat eggs at all. In fantail darters (Lindström and Sargent, 1997), the number of eggs eaten did not differ between total and partial cannibals. This would suggest that in certain species, a parent might consume a fixed amount of eggs. This also seems to apply to, e.g., the fathead minnow, Pimephales promelas (Sargent, 1988). However, the importance of a brood-size effect on full-clutch filial cannibalism has been confirmed by several studies (for sand gobies, see Forsgren et al., 1996; for common gobies, see Kvarnemo et al., 1998); i.e., small clutches were completely eaten more often than larger clutches, as predicted by Rohwer (1978) and Sargent (1992).
Energy cost
The amount of eggs eaten tended to affect a male's final energy state, so that those males that ate more of their eggs had more body fat at the end of the trial period. However, there was no correlation between a male's initial body condition and the amount of eggs eaten. Also, the body size of a male strongly tended to affect his final energy state, such that larger males had more body fat than did smaller males. Still, controlling for these effects, males subjected to the low oxygen treatment had a significantly lower amount of body fat than did males in the high oxygen treatment. Thus, rearing broods in a low dissolved oxygen environment clearly was more energy costly for males compared with rearing broods under fully saturated conditions. Similar results have been found in other studies as well. For instance, fantailed darter males who cannibalized their own eggs had better body condition at the end of the experiment (Lindström and Sargent, 1997), and cannibalistic sand goby males had higher fat reserves (Lindström, 1998). Further, common gobies with poorly built nests later consumed significantly more of their own eggs than did males with properly built nests, indicating that nest building conveys information about a male's energetic status (Kvarnemo et al., 1998). This means that in addition to the functional aspect of well-built nests, they may also act as a quality-revealing signal for females, much in the same way as an ornament does. This has been shown to be the case in a study on three-spined sticklebacks by Barber et al. (2001). Further, in the river bullhead, nest and egg guarding males lost a considerable amount of body mass (average mass loss was 13.5% and 18.8% in two populations) compared with that of other individuals of the same populations (Marconato et al., 1993). Male bluegill sunfish, Lepomis macrochirus, with larger broods lost more nonpolar lipids as a result of greater fanning expenditure (Coleman and Fisher, 1991), and parental male three-spined sticklebacks had significantly higher energy expenditure than did nonparental males (Smith and Wootton, 1999).
Nest building and nest entrance size
Nest entrance size and nest rebuilding was highly affected by treatment. Initially, there was no difference between the males assigned to the low oxygen treatment and males assigned to the high oxygen treatment, neither in nest building activities nor in nest entrance size. However, only three males in the low oxygen treatment rebuilt their nests after the treatment started, compared with 13 in the high oxygen treatment. In addition, nest entrance size was significantly larger in the low oxygen treatment group. This result is consistent with the studies of Jones and Reynolds (1999a,b,c) on male common gobies breeding in a low oxygen environment. Their studies also indicate that males may adjust their nest to a level that meets their ability to fan. An explanation would be that increased nest entrance area enhances fanning efficiency, as proposed by Jones and Reynolds (1999a,b,c) and Kvarnemo (1995), in which case a large nest entrance would represent an adaptive adjustment. Therefore, because nests with large nest openings were more likely to be discovered and attacked by shore crabs in the study by Jones and Reynolds (1999b), there may be a trade off between fanning efficiency and decreased nest predation. On the other hand, an uncovered nest does not necessarily have a larger entrance area, although they may be correlated. The increased bird predation that an uncovered nest may suffer from (Lindström and Ranta, 1992) suggests that there is a trade-off between nest camouflage and energy spent on nest building activities. Therefore, rebuilding the nest would mean an extra effort in addition to the increased cost of fanning. In accordance to this, Lindström (1998) found that male sand gobies with low fat reserves refrained from building nests, and Kvarnemo et al. (1998) showed that male common gobies on low diet were less likely to cover their nests.
On the other hand, results from Jones and Reynolds (1999b) suggest another possible explanation. They found that in normal oxygen conditions common goby females prefer to spawn in nests with a thick sand cover to nests that were not covered. However, this preference disappeared in low oxygen conditions, in which more fanning is required of the male. Consequently, if there is no female preference for covered nest in low oxygen environment, the males should have less incentive to allocate energy into building those nests. Further, Kvarnemo (1995) found that male sand gobies showed a preference for nests proportionate to their own size instead of the largest nest size available. Presumably, it is more important to choose a nest size that they will be able to fan and defend successfully against egg predators than to have as large nest area as possible for egg deposition.
Parental allocations and trade-offs
Despite the obviously greater amount of parental expenditure, in terms of fanning, made by males brooding in a low dissolved oxygen environment, we noticed no difference in the way these males performed their parental duties compared with that of males brooding under fully saturated conditions. In other words, there was no sign of dead eggs or any difference in appearance between clutches in replicates ran in the beginning of the reproductive season. Any dead or diseased eggs would probably have been removed, in which case filial cannibalism would have been considerably higher in the low oxygen treatment.
Later in the season though, we observed that two clutches reared by males in the control group had reached a more advanced developmental stage, the black eyes stage, than did two clutches reared by males in the experimental group during exactly the same trial period. This result is concordant with the findings of Jones and Reynolds (1999a), who observed a 1-day delay in hatching time for common goby clutches reared in a low dissolved oxygen environment. Developmental rate in sand goby eggs is strongly correlated with temperature, eggs developing faster with higher water temperatures (Fonds and van Buurt, 1974; Kvarnemo, 1994), and hence, a difference in developmental rate would only be noticeable later in the season when the water is slightly warmer. However, although hatching was delayed in the study of Jones and Reynolds (1999a), those males that completed care managed to achieve the same level of hatching success as the ones brooding in saturated waters. This indicates that the experimental males of their study, as well as ours, compensated for the increased need of ventilation by allocating more effort to fanning.
Taken into consideration that under these circumstances, the male himself must allocate more energy to breathing in order to maintain the same oxygen supply to tissues not involved in oxygen uptake, the extra energy spent on fanning should have even stronger implications (Kramer, 1987). The male would have less time and energy to spend on other activities such as foraging and nest maintenance, making the nest easier for predators to detect. Also, Jones and Reynolds (1999c) found that male common gobies reduced their nest ventilation in the presence of a predatory shore crab, Carcinus maenas, regardless of oxygen treatment. This indicates a trade-off between direct care and protection of eggs because an insufficient supply of oxygen could increase offspring mortality and cause a reduction in reproductive success.
According to Coleman et al. (1985) life-history models for parental investment assume that a parent is selected to maximize its remaining lifetime reproductive success according to a trade-off between present and future reproduction. Further, Carlisle (1982) developed models for parental care allocation to maximize brood success in variable environments. One of these models is the threshold model, to which sand gobies, among others, may apply. This model predicts that if the environment or the parent's physical condition deteriorates, the parent should increase the amount of care for its brood. We have shown that despite the higher energy expenditure, sand goby males breeding in a low oxygen environment did not eat more of their eggs than did males in a control treatment. This indicates that less energy will be available for later mating effort and the raising of future broods and, hence, represents an investment in present offspring survival, possibly at the expense of future reproductive success. Similarly, Jones and Reynolds (1999a) found that common goby males under low oxygen conditions were more prone to abandon their broods in the second brood cycle as a result of the greater fanning expenditure made during the first brood cycle. Further, in two studies made by Magnhagen (1990, 1993), sand goby males showed no change in nest building or spawning frequency in the presence of a predatory cod, whereas young black goby males, Gobius niger, would not build nests or spawn under these circumstances. The higher risk taking shown by both the sand goby and the closely related common goby probably reflects a difference in expected future reproductive success of these short-lived species compared with that of the longer living black goby.
To conclude, filial cannibalism in the male sand goby did not increase as a result of the experimental increase in parental effort. Instead, the extra effort made by males breeding in a low oxygen environment could be noticed by a significantly lower amount of body fat compared with that of males in the control treatment. Furthermore, the treatment had a significant effect on nest building. Thus, environmental factors, such as low levels of dissolved oxygen, have implications for allocation of parental effort and may affect lifetime reproductive success in the sand goby, both owing to limited energy recourse for future mating effort and because of a possible trade-off between direct care and nest defense.
|
ACKNOWLEDGEMENTS |
|---|
We wish to express our gratitude towards the following people: the staff of Tjärnö Marine Biological Laboratory, especially Bertil Rex and Karlanders Hagsköld for lending us equipment and advice on how to use them, Christer Wiklund for comments on the manuscript, and Andreas Nyman for sharing trawling, good times, and bad times at Tjärnö.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Baden S, Loo L-O, Pihl L, Rosenberg R, 1990. Effects of eutrophication on benthic communities including fish: Swedish west coast. Ambio 19:113-122.
Barber I, Nairn D, Huntingford FA, 2001. Nests as ornaments: revealing construction by male sticklebacks. Behav Ecol 12:390-396.
Belles-Isles J-C, FitzGerald GJ, 1991. Filial cannibalism in sticklebacks: a reproductive management strategy? Ethol Ecol Evol 3:49-62.
Breitburg DL, 1992. Episodic hypoxia in Chesapeake Bay: interacting effects of recruitment, behavior, and physical disturbance. Ecol Monogr 62:525-546.
Carlisle TR, 1982. Brood success in variable environments: implications for parental care allocation. Anim Behav 30:824-836.[CrossRef]
Clutton-Brock TH, 1991. The evolution of parental care. Princeton, New Jersey: Princeton University Press.
Coleman RM, Fisher RU, 1991. Brood size, male fanning effort and the energetics of a non-shareable parental investment in bluegill sunfish, Lepomis macrochirus (Teleostei: Centrarchidae). Ethology 87:177-188.
Coleman RM, Gross MR, Sargent RC, 1985. Parental investment decision rules: a test in bluegill sunfish. Behav Ecol Sociobiol 18:59-66.
FitzGerald GJ, van Havre N, 1987. The adaptive significance of cannibalism in sticklebacks (Gasterosteidae: Pisces). Behav Ecol Sociobiol 20:125-128.
Fonds M, Van Buurt G, 1974. The influence of temperature and salinity on development and survival of goby eggs (Pisces, Gobiidae). Hydrobiol Bull 8:110-116.[CrossRef]
Forsgren E, Karlsson A, Kvarnemo C, 1996. Female sand gobies gain direct benefits by choosing males with eggs in their nests. Behav Ecol Sociobiol 39:91-96.[CrossRef]
Gross MR, Shine R, 1981. Parental care and mode of fertilization in ectothermic vertebrates. Evolution 35:775-793.[CrossRef]
Guitel F, 1892. Observasions sur les moeurs du Gobius minutus. Arch Zool Exp Génér 10:499-555.
Hoelzer GA, 1992. The ecology and evolution of partial-clutch cannibalism by paternal Cortez damselfish. Oikos 65:113-120.
Jones JC, Reynolds JD, 1999a. Costs of egg ventilation for male common gobies breeding in conditions of low dissolved oxygen. Anim Behav 57:181-188.[CrossRef][ISI][Medline]
Jones JC, Reynolds JD, 1999b. The influence of oxygen stress on female choice for male nest structure in the common goby. Anim Behav 57:189-196.[CrossRef][ISI][Medline]
Jones JC, Reynolds JD, 1999c. Oxygen and the trade-off between egg ventilation and brood protection in the common goby. Behaviour 136:819-832.[CrossRef]
Kautsky N, Kautsky H, Kautsky U, Waern M, 1986. Decreased depth perception of Fucus vesiculosus (L.) since 1940s indicate eutrophication of the Baltic Sea. Mar Ecol Prog 28:1-8.
Kodric-Brown A, 1983. Determinants of male reproductive success in pupfish (Cyprinodon pecosensis). Anim Behav 31:128-137.[CrossRef]
Kondoh M, Okuda N, 2002. Mate availability influences filial cannibalism in fish with paternal care. Anim Behav 63:227-233.[CrossRef]
Kraak SBM, Weissing FJ, 1996. Female preference for nests with many eggs: a cost-benefit analysis of female choice in fish with paternal care. Behav Ecol 7:353-361.
Kramer DL, 1987. Dissolved oxygen and fish behaviour. Env Biol Fishes 18:81-92.
Kvarnemo C, 1994. Temperature differentially affects male and female reproductive rates in the sand goby: consequences for operational sex ratio. Proc R Soc Lond B 256:151-156.
Kvarnemo C, 1995. Size-assortative nest choice in the absence of competition in males of the sand goby, Pomatoschistus minutus. Env Biol Fishes 43:233-239.
Kvarnemo C, Svensson O, Forsgren E, 1998. Parental behaviour in relation to food availability in the common goby. Anim Behav 56:1285-1290.[CrossRef][ISI][Medline]
Lindström K, 1998. Effects of costs and benefits of brood care on filial cannibalism in the sand goby. Behav Ecol Sociobiol 42:101-106.
Lindström K, Ranta E, 1992. Predation by birds affects population structure in breeding sand goby, Pomatoschistus minutus, males. Oikos 64:527-532.[CrossRef][ISI]
Lindström K, Sargent RC, 1997. Food access, brood size and filial cannibalism in the fantail darter, Etheostoma flabellare. Behav Ecol Sociobiol 40:107-110.
Magnhagen C, 1990. Reproduction under predation risk in the sand goby, Pomatoschistus minutus, and the black goby, Gobus niger: the effect of age and longevity. Behav Ecol Sociobiol 26:331-335.
Magnhagen C, 1993. Conflicting demands in gobies: when to eat, reproduce, and avoid predators. Marine Behav Physiol 23:79-90.
Manica A, 2002. Filial cannibalism in teleost fish. Biol Rev 77:261-277.[Medline]
Marconato A, Bisazza A, Fabris M, 1993. The cost of parental care and egg cannibalism in the river bullhead, Cottus gobio L. (Pisces, Cottidae). Behav Ecol Sociobiol 32:229-237.
Miller PJ, 1986. Fishes of the north-eastern Atlantic and the Mediterranean, vol. 3. Paris: Unesco.
Perrin N, 1995. Signalling, mating success and paternal investment in sticklebacks (Gasterosteus aculeatus): a theoretical model. Behaviour 132:1037-1057.
Pihl L, Magnusson G, Isaksson I, Wallentinus I, 1996. Distribution of ephemeral macroalgae in shallow bays on the Swedish west coast. J Sea Res 35:169-180.[CrossRef]
Pihl L, Svenson A, Moksnes P-O, Wennhage H, 1999. Distribution of green algal mats throughout shallow soft bottoms of the Swedish Skagerrak archipelago in relation to nutrient sources and wave exposure. J Sea Res 41:281-294.[CrossRef]
Petersen JK, Petersen GI, 1990. Tolerance, behaviour and oxygen consumption in the sand goby, Pomatoschistus minutus (Pallas), exposed to hypoxia. J Fish Biol 37:921-933.[CrossRef]
Rohwer S, 1978. Parent cannibalism of offspring and egg raiding as a courtship strategy. Am Nat 112:429-440.[CrossRef][ISI]
Sargent RC, 1988. Parental care and egg survival both increase with clutch size in the fathead minnow, Pimephales promelas. Behav Ecol Sociobiol 23:33-37.[CrossRef]
Sargent RC, 1992. Ecology of filial cannibalism in fish: theoretical perspectives. In: Cannibalism: ecology and evolution among diverse taxa (Elgar MA, Crespi BJ, eds). New York: Oxford University Press; 38–62.
Sikkel PC, 1989. Egg presence and developmental stage influence spawning-site choice by female garibaldi. Anim Behav 38:447-456.[CrossRef]
Smith C, Wootton RJ, 1999. Parental energy expenditure of the male three-spined stickleback. J Fish Biol 54:1132-1136.[CrossRef]
Stearns SC, 1989. Trade-offs in life-history evolution. Funct Ecol 3:259-268.[CrossRef]
Svensson O, Magnhagen C, Forsgren E, Kvarnemo C, 1998. Parental behaviour in relation to the occurrence of sneaking in the common goby. Anim Behav 56:175-179.[CrossRef][ISI][Medline]
Takegaki T, Nakazono A, 1999. Responses of the egg-tending gobiid fish Valenciennea longipinnis to the fluctuation of dissolved oxygen in the burrow. Bull Mar Sci 65:815-823.
Torricelli P, Lugli M, Gandolfi G, 1985. A quantitative analysis of the fanning activity in the male Padogobius martensi (Pisces: Gobiidae). Behaviour 92:288-301.
Zoran MJ, Ward JA, 1983. Parental egg care behaviour and fanning activity for the orange chromide, Etroplus maculatus. Env Biol Fish 8:301-310.[CrossRef]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  O. Svensson and C. Kvarnemo Parasitic spawning in sand gobies: an experimental assessment of nest-opening size, sneaker male cues, paternity, and filial cannibalism Behav. Ecol., March 1, 2007; 18(2): 410 - 419. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Kvarnemo Evolution and maintenance of male care: is increased paternity a neglected benefit of care? Behav. Ecol., January 1, 2006; 17(1): 144 - 148. [Full Text] [PDF]
|
|
|
|
|
|
O. Svensson and C. Kvarnemo The importance of sperm competition risk and nest appearance for male behavior and female choice in the sand goby, Pomatoschistus minutus Behav. Ecol., November 1, 2005; 16(6): 1042 - 1048. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||