Role of settlement in determining the distribution and abundance of barnacles in a temperate mangrove forest

Abstract

An increased understanding of the role of early life history stages in determining the distributions of benthic invertebrates require studies covering a wide range of habitats, and preferably multiple scales. For one of the more prominent invertebrate groups, intertidal barnacles, much of our existing knowledge is drawn from studies of rocky shores, with few studies from other habitats. Here, we describe the distribution of adults of the temperate barnacle Elminius covertus in mangrove (Avicennia marina) forests of southeastern Australia. We sampled E. covertus on pneumatophores over a large-scale (50–100 m) across the forests from the landward to seaward edges. We also described large scale patterns along the shore (also a scale of 100 m). At the seaward edge of the forest, on a small spatial scale, the vertical distribution of E. covertus on pneumatophores over 15 cm was also investigated. Settlement and recruitment were also estimated by allowing barnacles to settle on artificial substrata. The processes influencing the patterns of adults of E. covertus were very different over the two major scales in this study. At the large-scale, adult Elminius covertus were abundant at the seaward edges of forests, declining through the forest, and were absent at the landward sections of forests. Recruitment, measured over 1 month, and 1 week, matched this pattern, and settlement, measured over a single high tide, also matched the adult distributions. Along the shore, differences in abundance of adult barnacles were matched by variation in settlement and recruitment. Post-settlement mortality had little influence on this pattern. In contrast, on the small, vertical scale, barnacles were most abundant on the upper 5 cm of pneumatophores, and least common near the substratum. This vertical pattern reflected the pattern of recruitment after 1 month. Settlement, however, was different, with more settlement on the 5 cm of pneumatophores closest to the substratum. The distribution of recruits after 1 week was intermediate between settlement and recruitment after 1 month. The vertical pattern of adults is therefore determined by post-settlement mortality occurring during the first month after settlement.

Introduction

Recent ecological research has demonstrated that variability in larval settlement is also a strong determinant in shaping population structure (Denley and Underwood, 1979, Grosberg, 1982, Yoshioka, 1982, Keough, 1984, Underwood and Denley, 1984, Connell, 1985, Gaines and Roughgarden, 1985, Gaines et al., 1985, Roughgarden et al., 1985, Lewin, 1986, Underwood and Fairweather, 1989, Raimondi, 1990, Sale, 1990, Minchinton and Scheibling, 1991).

Despite increased attention to settlement and recruitment processes, little generality has emerged. Overall levels of settlement have been suggested as a factor that may determine the relative role of settlement, with populations more likely to be settlement-limited at low settlement rates (Connell, 1985). In a similar way, regional productivity, as reflected by regional nutrient levels in the water column, may influence larval production, and hence the role of settlement or recruitment in population dynamics (see, e.g. Menge et al., 1997 for an example linking productivity and community structure).

We see three important obstacles to emergent generality about the role of settlement and/or recruitment: the number of studies focusing on a restricted number of life history stages, the representativeness of the database, and potential scale-dependence of any results.

The first limitation is easily remedied — questions about the importance of recruitment can be answered meaningfully only if they involve information about other life history stages (Keough, 1988). Because recruitment is an anthropocentric, rather than an objective biological term, it is also important to separate settlement from recruitment, to determine whether settlement or post-settlement processes are critical (Keough and Downes, 1982).

Secondly, the database from which to draw any general inferences about invertebrate recruitment is strongly biased towards the intertidal zone of rocky shores. For example, in barnacles, one of the most prominent (and over-represented) groups in the literature, most studies examining the distribution and abundance have been undertaken on rocky intertidal areas (Moore, 1935, Connell, 1961a, Connell, 1961b, Achituv, 1972, Denley and Underwood, 1979, Wethey, 1983, Wethey, 1984, Wethey, 1985a, Wethey, 1985b, Caffey, 1985, Otaiza, 1989, Sutherland, 1990, Raimondi, 1990, and many others). Barnacles are also common elsewhere (e.g. in mangrove forests), but there have been few studies examining the processes that determine the distribution and abundance of barnacles living in these areas (Bayliss, 1993, Coates and McKillup, 1995, Ross and Underwood, 1997).

Thirdly, the relative importance of settlement, early post-settlement, and adult processes may change with the scale under consideration. For example, Keough and Chernoff (1987) attributed large-scale patchiness in the spatial distribution of the bryozoan Bugula neritina to accidents of dispersal, and not to post-settlement processes. In contrast, Keough (1986) found that, at smaller scales, along seagrass leaves, distributional patterns were explained by settlement behaviour and consistent gradients in post-settlement (>1 week after settlement) mortality and growth over scales of 10 to 20 cm.

Here, we examine the importance of settlement in determining the distribution of a barnacle, Elminius covertus, living in temperate mangrove forests of southeastern Australia. We describe the patterns of distribution of Elminius covertus at two spatial scales, horizontally (cross-shore) in a mangrove forest, and, on a smaller scale, vertically from 0 to 15 cm above the sediment surface, on mangrove pneumatophores. We also generated information about barnacle distributions at a third scale, along the shore. By measuring adult distributions, recruitment (using two different temporal windows), and settlement, we determined whether post-settlement mortality influences these patterns of distribution.