The ecology and reproductive biology of Ericentrus rubrus (Clinidae), with an assessment of rockpool fish community structure

Abstract

Ericentrus rubrus (Hutton, 1872) is the commonest and most widespread of the three clinid species in New Zealand waters. A regular inhabitant of intertidal rockpools (although also occurring subtidally), E. rubrus was sampled from this habitat on a monthly basis from June 1991 until August 1993, and a study conducted on the reproductive biology and population structure of the species. Meristic comparisons of E. rubrus and another clinid, Cologrammus flavescens (Hutton, 1872) made from preserved museum specimens taken from different locations showed no significant variation over different latitudes or between the mainland and the Chatham Islands. The placing of the two species in separate genera is confirmed. E. rubrus is viviparous, with superfoetation (simultaneous development of multiple stages of eggs and/or embryos). Gestation is intrafollicular, and the mode of embryonic nutrition shifts from lecithotrophy to trophodermal matrotrophy and, finally, intestinal nutrient transfer in a similar manner to the Australian and South African Clinidae. Maternal investment in individual offspring is estimated (conservatively) to be 2640%. Fecundity is very low for a teleost, fewer than 200 progeny being produced each year from eight to ten broods, and is exponentially related to maternal size. Parturition is asynchronous, and occurs over several months in spring and summer, following which the larvae spend several weeks in the plankton close to littoral surge channels. Post-settlement growth is rapid. Juveniles reach maturity in the summer following parturition, and most of the reproductive population consists of individuals in the 1+ year class. E. rubrus lives to a maximum of four years, by which time it has reached a total length of 105mm. The sex ratio is approximately 1:1, and there is no sexual dichromatism or dimorphism, other than the presence of an intromittent organ in the male and when females are conspicuously pregnant. An experimental approach to the community ecology of the Wellington rockpool fish assemblage is presented. Repeated defaunation of pools over a one year period allowed natural seasonal variation and the effects of sampling-related disturbance to be determined. This community was found to be resilient, persistent and stable, with recolonisation usually occurring within two months of defaunation. Rockpools are apparently not important nursery areas for subtidal species, but are important refugia for juveniles of habitually intertidal fishes. The timing of recruitment for various species agrees with that recorded in previous studies. Analytical methods for the detection of disturbance in community structure, which had not been previously used for intertidal fish communities were applied. Recommendations for future study are made concerning fish sampling methodology, the application of abundance/biomass comparisons and the transformation of data for non-parametric multivariate analysis of shore fish communities. A new method of assessing intertidal fish density is proposed, that of the number of fish per unit pool substrate area. E. rubrus is a poor recoloniser and decreases in abundance in the intertidal during the winter. This suggests that the species is strongly K-selected and its presence in large numbers may be indicative of habitat stability.