Constancy of Mean Size in Starfish


Chapter 8. Constancy of Mean Size

8.1 Introduction

The fluctuations that occur in animal populations have been regarded as a measure of community stability (MacArthur, 1955; Frank, 1968; Den Boer, 1971; Jacobs, 1974; Goodman, 1975; Brown, 1981). Many authors believe that complex high diversity systems are characterised by relative constancy of species composition (e.g. Dunbar, 1960; 1972; Leigh, 1965; Margalef, 1963; 1974). They propose that populations of the component species do not vary to the extent demonstrated in more simple communities. Additionally, the interaction of competitors and predator / prey situations might prohibit the resource monopolisation so characteristic of dominant species in less diverse systems. Other authors (e.g. Connell, 1978; Sale, 1976, 1977; Sale and Douglas, 1984) believe that there is temporal variability in the community structure of the coral-reef organisms they have studied.

A paucity of juveniles characterises the population structures of large bodied, coral-reef starfish (Yamaguchi, 1973 a). It is possible that populations are maintained either by continual low recruitment or occasional high recruitment, each coupled with iteroparity. The juveniles are cryptic and their apparent absence or rarity indicates that reproductive success is either constantly low, sporadic or both. Amongst coral-reef species, population outbreaks have been well documented for only Acanthaster planci. However, population changes in Linckia laevigata following Acanthaster outbreaks have been suggested by Laxton (1974) and Asterina burtoni is known to have extended its range into the Mediterranean Sea following its introduction through the Suez canal (Achituv, 1969).

In most common species of coral-reef asteroid that have been studied, their reproductive strategy was directed towards the production of enormous quantities of gametes which were released directly into the surrounding water (Yamaguchi, 1973 a; 1977 a). If the mortality of the resulting larvae varied greatly from year to year, then we would expect years of noticeable recruitment followed by one or more years of little or no recruitment success. If the adult population is short-lived (e.g. two years), then the recruitment necessary to maintain the population must occur within this short period. If there were another consecutive year of recruitment failure then the species would become locally extinct. In these short-lived species, juveniles should occur in sufficient numbers to be detected. If, however, the adults are long-lived, then the level of recruitment required to maintain the adult population could be extremely low per annum and we might expect to see juveniles only occasionally.

If periods of high recruitment are required to maintain the population structure of a common species, the mean individual size of the populations should vary as a consequence of the influx of juveniles. In Ophidiaster granifer, when periods of recruitment occurred at Guam, the mean individual size of the population decreased (see Yamaguchi and Lucas, 1984). The mean individual size should increase progressively throughout the interval between periods of recruitment. In a large bodied species, such as Linckia laevigata, the mean individual size should increase slowly (dependent on growth rate), but might reach a size equilibrium determined by the availability of food (Paine, 1976). If periods of recruitment occurred, the mean individual size of the population should decrease. If recruitment did not occur, the mean individual size should increase slowly. The rate of increase in mean individual size will be determined by the average individual growth rate. This might be very slow in a species that is long lived.

8.2 Methods

The size data for the more common species were analysed to see if there was any temporal variation in mean size for each species. A one way ANOVA (ratio of variance in mean major radius (R mm) between and within sampling periods) for each species was calculated and the results are listed in Tables 5.1 – 5.10.

The mean size variation required to produce a probability level of .05 or .01 was very small for these relatively common species. Any probability not less than 0.001 represented only a small mean size variation compared with the size variation within each of the populations.

8.3 Results

Table 8.1 The significance of temporal size variation for each of the more common species. The grand mean (R mm), sample size (N), F statistic, degrees of freedom and probability level of the size variation over several sampling periods are tabled. See Figure 8

SPECIES                    MEAN R     N      F    d.f.    P
Linckia guildingii           134     131    1.1   9,82     N/S
Linckia laevigata            127     516    1.7  11,459   <.05
Linckia multifora             38     396    9.5  10,329  <.001
Nardoa novaecaledoniae        88     361    3.2  11,295   <.01
Nardoa pauciforis            104     233    2.2  11,179   <.05
Disasterina abnormalis        15    1109   15.8 12,1054  <.001
Echinaster luzonicus          48     988   11.4 10,883   <.001 

8.4 Discussion

From Tables 5.1 – 5.10, Figures 5.1a – 5.10a and Table 8.1, it can be seen that four of the large-bodied species, Linckia guildingii, Linckia laevigata, Nardoa pauciforis and Nardoa novaecaledoniae, did not vary their mean size greatly over the entire study period of five years. In two of these species, Linckia laevigata and Nardoa pauciforis, the mean size variation was significant at 0.05. In Nardoa novaecaledoniae, while significant at 0.01, this variation still represented only a small change in the mean size of the population over the entire study period.

Although four of the seven common species maintained a size distribution that did not vary greatly during the study period, the possibility that many of the species might demonstrate occasional high recruitment success, when observed on a much larger time scale, cannot be rejected. If such recruitment occurred, it would manifest itself as oscillations in the mean individual size of that species. Asteroids are also known to possess highly plastic growth rates which can effectively disguise annual year classes.

It should be noted that a stable size distribution does not necessarily imply low recruitment and low mortality, but can result from a balance of high recruitment and high mortality. Under conditions of high mortality and low recruitment, a population with a low growth rate can also manifest a stable size distribution but it would show a simultaneous decline in population density. This was not observed in the present study and the small change in the mean size of the large-bodied species suggests that Linckia guildingii, Linckia laevigata, Nardoa novaecaledoniae and Nardoa pauciforis are long-lived.

During this period, Linckia multifora, Disasterina abnormalis and Echinaster luzonicus showed mean size variations that were highly significant. This size variation was the result of periodicity in either sexual or asexual reproduction. In Linckia multifora and Echinaster luzonicus, the difference in size resulted from autotomy. High recruitment of juveniles was observed in only one small bodied, sexually reproducing species, Disasterina abnormalis. In the remaining species, the abundances were low, and statistically valid comparisons of what might have been temporal mean size variation could not be justified. This applied to Ophidiaster granifer that showed periodic recruitment when studied at Guam (Yamaguchi and Lucas, 1984), and Asterina burtoni which did not show significant mean size variation in the present study.

The relative stability of the size distributions of the common large-bodied species can be explained by assuming very slow growth of a predominant year class or a balance of recruitment and mortality within each of the species. It seems likely that a combination of both is involved. The paucity of juvenile asteroids, and the constancy of the size distributions in all the large bodied sexually reproducing species can be explained only by a life-history model which incorporates low adult mortality and includes the assumption of longevity.

The variation in mean size between populations of Linckia laevigata at different localities in the Indo-West Pacific could be caused by the presence of geographically asynchronous, dominant year classes. However, this is unlikely as this species did not alter its mean size greatly during the period of the present study. The highly plastic growth rate may be influenced by nutrition (see Wolda, 1970) or other factors (e.g. disturbance) may cause both the higher density and smaller mean size. Dwarfism, resulting from high salinity, was described in Asterina burtoni by Price (1982).

The results of this study of coral-reef asteroids contrasts with data relating to laboratory rearing of Acanthaster planci which are claimed to demonstrate individual senescence at an age of approximately five years (Lucas, 1984). This finding, which appears to be inconsistent with the general biology of an often rare, large-bodied and venomous animal, can be attributable to the laboratory rearing conditions (see Endean and Cameron, 1990 b). Additionally, a specimen of Acanthaster planci held in an aquarium at the Heron Island Research Station decreased to two-thirds of its original size within a period of 6 months. When adequate food is not available, regression in size might occur in many coral-reef asteroid species. At Heron Reef, the coral-reef asteroid community is not dominated by violently fluctuating size structures as might be expected from the work of Lucas (1984). All the large-bodied, sexually reproducing asteroids in this study existed with a stable size structure for the entire study period.


One thought on “Constancy of Mean Size in Starfish

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s