Temporal variation in richness and composition of recruits in a diverse cnidarian assemblage of subtropical Brazil

Highlights

• Cnidarian richness occurs in a temporally bimodal pattern in this assemblage.

• Local richness may be better estimated by increasing sampling efforts over time.

• Time of substrate exposure did not influence species richness, but composition.

• Microhabitat variation was not important for species richness.

Abstract

Many studies have shown that the processes of colonization of new marine substrata, from settlement to recruitment, can leave long and lasting signals on the richness, composition and general structure of natural epifaunal assemblages. Systematic descriptions of temporal variability in patterns of richness and structure of recruits are scarce, partly because of logistical difficulties of working with multispecies assemblages of recruits. Here we quantify temporal variation in recruit richness, composition and structure of a rich cnidarian assemblage in southeastern Brazil, and evaluate the effect of microhabitat type and time of submersion on these patterns. We conclude that hydrozoan (the prevailing cnidarians in this assemblage) species richness occurs in a temporally bimodal pattern, with the majority of species divided between year-round recruiting species and temporally infrequently, non-seasonal species. This pattern does not depend on the species local abundance, and suggests that local species richness may be better estimated by increasing sampling efforts over time. Moreover, time since substrate submersion had no effect on species richness, but influences species composition, suggesting constant changes in the assemblage instead of accumulation of species through time. Finally, microhabitat variation, measured as differences between sheltered and exposed surfaces, despite influencing species' abundances, was not important for species richness.

Introduction

One of the primary determinants of the local number of species observed in marine epifaunal communities is the colonization and early survival of propagules. These will be a function of the species found within dispersal distance of the new surface that becomes available (Navarrete, 2007, Palardy and Witman, 2011, Palardy and Witman, 2014). Species richness can and will be modulated by species interactions and differential mortality of later stages, but successful colonization, i.e. recruitment, is required to become part of the local assemblage and, in many systems, this recruitment process can leave a long-lasting signal in the diversity and relative abundance of species of the adult epifaunal community (Bertness et al., 2001, Caro et al., 2010, Hubbell, 1997, Palardy and Witman, 2011, Palardy and Witman, 2014, Sutherland, 1981). Since propagules of all species are not equally abundant over space and time (e.g. Benedetti-Cecchi et al., 2003, Crowder and Figueira, 2006, Nandakumar, 1996, Navarrete et al., 2008, Santelices, 1990, Sutherland and Karlson, 1977), it is of great importance to characterize and better understand the processes underling variability in species richness and composition of recruiting propagules and its influence on adult community structure in marine systems.

High temporal variability in the species composition, richness and overall arrival rates of new recruits to benthic communities characterizes most benthic epifaunal communities (Caffey, 1985, Caro et al., 2010, Palardy and Witman, 2011, Sutherland, 1981, Watson et al., 2011). One of the explanations for these temporal changes is variation in sea surface temperature (Clarke, 2009), which interacts with annual cycles of adult reproduction and larval release, causing variability in larval availability of different species at any given time (Coma et al., 2000, Cowen and Sponaugle, 2009, Flores and Negreiros-Fransozo, 1998). Temporal variation in richness and composition of recruits can also be driven by variation in patterns of circulation, larval dispersal and cross-shelf transport to shore since most of these mechanisms differentially affect larvae of different species (e.g. Narváez et al., 2006, Watson et al., 2011). Changes in turbulence and local flows over scales of centimeters to few meters can also determine whether larvae of a given species can actually reach the settling surface, with consequent effects on richness and composition of the settler assemblage (Palardy and Witman, 2014, Pineda et al., 2010). Moreover, since settlement in most invertebrate species is the result of active larval selection and the existence of chemical cues from previous settlement events and biofilms (Butman, 1987, Hadfield, 2011, Jenkins, 2005, Keough and Downes, 1982), temporal variation in these chemical signals can determine which species successfully colonize a surface. All of these processes can cause temporal variation in the structure and richness of the settling larval assemblage.

After settlement, differential early mortality will further modify the structure of the recruit community found on the surface (Gosselin and Qian, 1997). Therefore, factors affecting early mortality (e.g. substrate heterogeneity, substrate temperature) can further modulate the structure of the recruit assemblage. As time after settlement increases, the more important these pos-settlement factors of mortality become. Indeed, studies on artificial panels find that community development varies relative to both time of submersion (Bram et al., 2005) and colonization success (Keough and Downes, 1982). Interactions between different taxa comprising the local community may have consequences from larval settlement to adult survival (Osman and Whitlatch, 1995, Osman et al., 1989). Competition for space may limit the arrival of new recruits to an established community (Jackson, 1977), while concurrently, post mortality may free space for settlement by new species which may gradually replace existing ones (Minchinton and Scheibling, 1993). Thus, community composition and patterns of species abundance of the recruit and early juvenile stages of epifaunal communities is inexorably related to the time that the surface remains submerged. Here we evaluate whether there are characteristic patterns of temporal variation in richness, species composition and relative abundance of cnidarian species that recruit to highly diverse epifaunal communities of subtropical Brazil, and how these patterns change with time since the substrate was submerged and first colonized by new arrivals.

Cnidarians have complex life cycles that usually comprise two main stages: polyp (generally benthic) and medusa (generally planktonic), which sometimes include reduced, derived phases (Marques and Collins, 2004). Hydrozoan cnidarians are the commonest taxa on recruitment panels along Brazilian shores and opportunistic settlement (Calder, 1991b, Migotto et al., 2001) allows them to colonize a variety of habitats and substrates (Gili and Hughes, 1995). Hydrozoan have different modes of sexual (planula and actinula larvae) and asexual (buds, frustules or colony fragments) propagation (Gili and Hughes, 1995). Therefore, rapid larval settlement from the plankton and rapid asexual initial growth of colonies help explain why they quickly show on bare substrates and over other organisms (Boero, 1984, Migotto et al., 2001). It is also common to see some hydroids that develop into relatively large and robust colonies, which can then resist settlement and overgrowth by other sessile invertebrates, such as sponges, tunicates, and bryozoans (Migotto et al., 2001). Seasonal variation in abundance of hydroids has been associated with activity–quiescence cycles (Bavestrello et al., 2006), recruitment (Migotto et al., 2001), and reproduction (Gili and Hughes, 1995), and fluctuations often follow sudden changes in sea surface temperature (Calder, 1990, Migotto et al., 2001). Most studies on cnidarian seasonality have focused on one or few common species (Bavestrello et al., 2006, Calder, 1990, Migotto et al., 2001), while temporal variation in assemblage structure, including richness and relative commonness and rarity in time and space has not been amply reported.

Since no previous information exists in this system, our aim here is to evaluate simple hypotheses regarding richness and composition of cnidarian assemblages as they recruit to new surfaces and change over time. Using settling panels replaced every 3 mo we examined whether the total number of species observed after 2 years is primarily the result of different species recruiting over time (high temporal Beta diversity, Anderson et al., 2011), or whether a large fraction of all species is available at any given point in time. The latter means that there is a comparatively low species turnover in time and that studies could estimate cnidarian species richness for the region by increasing sample size a single time. It also implies relatively low seasonality in richness and recruitment of most species, but comparatively high species turnover from month to month. Comparisons with plates submerged throughout the year allowed us to test whether total richness is the result of species (passively) accumulating on the panels through the year, or whether significantly higher extinctions occurred as species settle and grow in the plates, as expected for instance, if competitive exclusion is important on the plates.

Since the spatial distribution of marine organisms is predicted to be bimodal, with abundant and widely distributed core species, and rare and patchily distributed satellite species (Hanski, 1982), we evaluate whether such a core–satellite pattern characterize the cnidarian recruit assemblage in terms of temporal persistence and spatial occupation, as it might be expected from the dependency of species distributions also on sampling time interval (Magurran, 2007). We hypothesized that 1) the long-term (yearly) and typically high hydrozoan species richness is the result of accumulation of species recruiting on different times of the year; 2) the structure (composition and abundance) of the assemblage changes over time; 3) microhabitat (sheltered or exposed plates) influence the cnidarian recruit assemblage structure; and 4) the structure of the cnidarian assemblage after one year converges to a more homogeneous configuration, different than all recruitment events, as observed for instance in rocky shore communities (Caro et al., 2010).