Small-scale variation in tropical tidepool fish assemblages

Abstract

Habitat heterogeneity is known to promote species diversity and other effects in communities by increasing structural complexity, access to food resources and to refuge spaces. However, the influence of environmental heterogeneity at small spatial scale settings is challenging to detect and to measure. In tropical intertidal rocky shores, the low tidal zones are sometimes encrusted by reef-building organisms forming complex biogenic flat reefs below rocky outcrops that appear at mid-to-upper intertidal level. Here, we investigated a tidepool community situated on an intertidal shore with marked variation in structure and texture within metres to test whether tidepool fauna would differ between biogenic and rocky reefs in regard to their composition, diversity and trophic structure. Our findings revealed that tidepool fish assemblages differed significantly between reef types. Trophic structure exhibited variations in resource use, food web hierarchy and trophic roles of species occurring in both biogenic and rocky tidepools. Our results imply that tropical intertidal reefs show pronounced taxonomic and trophic distinctiveness at small spatial scales.

1. Introduction

Spatial heterogeneity is a strong predictor of species distribution patterns and trophic interactions [1,2] because landscape and seascape variation across space at various scales is responsible for structuring communities ranging from large-bodied mammals to cryptic and minute organisms living within forest leaf debris or in interstices of the reef matrix at sea [3–5]. For marine ecosystems, ordinary ecological measures excel at indicating the effects of spatial heterogeneity across habitats [6,7]. However, within-habitat changes in community or ecological processes at the metre-scale are challenging to observe.

Rocky shores are ideal locations to examine shifting patterns of distribution of organisms along a habitat continuum as sharp environmental gradients, especially the vertical unidirectional stress gradient, often shape distributions into clear patterns of vertical zonation [8–10]. Universal zonation schemes are largely based on the distribution of flora and sessile (and slow-moving) animals on the emerged substrata [8,9]. In contrast, comparatively little research has been dedicated to tidepool (rockpool) biota. Some attention has been given to fish, however, with a number of studies on American, African, Asian, European and Oceanian tidepool fishes having been published in recent years [11–15].

Tidepools are natural mesocosms providing a model system for testing community and trophic theories [16,17]. Their characteristics, including partial isolation during low tide, well-defined boundaries and set size at full exposure, facilitate effective manipulation and sampling for a wide range of ecological investigations [18]. The vertical position of tidepools influences community structure and species distribution [19–21] despite the vertical zonation patterns of tidepool fauna being much less conspicuous than those observed on emergent intertidal substrata. Zonation is more apparent on steeply sloping hard rocks on temperate and sub-tropical shores. In contrast, softer carbonate-dominated tropical shores are typically gently sloping with some growth of reef-building organisms just above the low-water mark [22–25].

Reef-builder organisms, such as vermetid gastropods, corals and coralline algae, generally encrust intertidal reefs up to less than 1 m above the low intertidal mark height of the rocky intertidal platform depending on the tide height and wave exposure. Above this, the rocky platform is often bare or colonized by fewer algae and invertebrate species.

At the regional scale, substratum nature is an important driver for the structure of tidepool fish assemblages, with some species demonstrating affinities to particular lithologies [19,26]. Rock-type may even influence tidepool fish distribution at small scales, given the spectra of specialized physiological, morphological and behavioural adaptations of intertidal fishes [11,27]. Here, we compared the composition, diversity and trophic structure using stable isotopes of the tidepool biota between biogenic and rocky substrates within the same intertidal reef zone. It is hypothesized that composition, diversity and trophic structure differ significantly between reef types, as biogenic tide pools—located adjacent to the shore—exhibit higher habitat complexity compared with the relatively bare rocky tidepools.

2. Methods

(a). Study area

The study area was Portinho Sul beach, which is located at the Santa Barbara Island, Abrolhos Marine National Park, Brazil (17°56′ S; 38°41′ W; figure 1), approximately 60 km off the Brazilian central coast. In this region, the tidal range is approximately 2.2 m. Santa Barbara measures 1500 m in length, with its intertidal zone composed mainly of Eocene basaltic rock and, to a lesser extent, of Cretaceous sandstone and shale and of coralline-algae bioclasts [28]. A biogenic flat reef, predominantly composed of vermetid gastropods and coralline algae, occurs at a low intertidal level about 0.4–0.7 m above mean low-spring tide level [28]. Its age is estimated to be 908−748 years before present. Tidepools are in both the biogenic flat reef (low and mid intertidal) and in the rocky basaltic shore (high intertidal) (figure 1). As the tide rises, the entire intertidal zone is inundated within an hour. All tidepools (biogenic and rocky) are flooded during both neap and spring tides, and there are no supratidal tidepools in the Portinho Sul beach.

Figure 1.

(a) Abrolhos archipelago in the central coast of Brazil (red dot) and the sampling site ‘Portinho Sul’ (black dot) on the intertidal reef (green) of Santa Barbara Island; emerged land in grey. (b) Biogenic (light substrate) and rocky (dark substrate) intertidal reef of Portinho Sul. Close view of the (c) biogenic and (d) rocky intertidal reef at Portinho Sul. (Photos: H. Guabiroba and R. Andrades.)

(b). Field sampling

Twenty tidepools were sampled at low tide in August 2019, with 10 in the biogenic reef and 10 in the rocky reef. The selection of tidepools for sampling was based on criteria that included the absence of connectivity to the sea or to other pools and similarity in size. The tidepools sampled were relatively small and shallow, with mean (± s.e.) depths of 5.1 (0.4) and 8.0 (± 1.7) cm in biogenic and rocky substratum, respectively, and volumes of 39.0 (± 7.8) and 58.9 (± 26.7) l, respectively.

Fish were collected using hand and dip nets after the application of eugenol (40 mg l⁻¹) in the 20 selected tidepools. Fish species were classified in relation to their degree of residency in tidepools, which are primary residents, secondary residents and transients. Primary residents are those species that spend their whole lifetime in the intertidal, while secondary residents (also known as partial residents or opportunists) refer to species that spend only part of their lifetime in the intertidal, usually as juveniles. Transients are species that use intertidal areas for feeding or refuge during short-time periods (e.g. tidal cycle), or simply pass through as they move between coastal habitats [15].

Concurrently with the fish collection, invertebrate consumers and primary producers were collected in other biogenic and rocky pools. This approach was employed to integrate the trophic structure analysis of the community based on stable isotopes. Crustaceans, polychaetes and gastropods were collected manually. Sample sizes for fishes and invertebrates are detailed in electronic supplementary material, table S1. Benthic and pelagic primary producers were collected in both reef types. Benthic producers (macroalgae) were collected using a scraper. Pelagic primary producers, phytoplankton (the principal source of particulate organic matter—POM), were extracted from water samples. These were sieved through a 60 μm mesh to remove the largest particles and zooplankton, and then filtered through pre-combusted Whatman GF/F glass-fibre filters. Each filter was then stored in a clean plastic vial prior to laboratory analysis. All samples were kept frozen at −20°C.

In the laboratory, muscle samples were extracted from the anterolateral portion of the trunk in fishes, from chelipeds in crabs and from soft tissues, avoiding shell and other carbonate structures in gastropods. Amphipods, isopods and polychaetes were wholly used for isotopic analysis after the removal of their gut content. All samples were dried at 60°C in a standard laboratory oven and then homogenized with a mortar and pestle. Samples were analysed for stable carbon and nitrogen isotopes using an elemental analyser (Flash, 2000 - Thermo Scientific, Milan, Italy) coupled to an isotope ratio mass spectrometer (Delta V Plus with a Conflo IV interface, Thermo Scientific, Bremen, Germany). The laboratory data were calibrated using reference materials (USGS-24, IAEA-CH6 and IAEA-600 for carbon and IAEA-N2, IAEA-NO-3, IAEA-600 for nitrogen). The analytical precision of the measurements was <0.15‰ for carbon and nitrogen based on internal standards. Stable isotope values are expressed using δ (delta) notation and parts per thousand (‰) as follows:

$${\displaystyle \delta \mathit{X}=\left[\left(\frac{R_{\mathrm{s}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{e}}}{R_{\mathrm{s}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{a}\mathrm{r}\mathrm{d}}}\right)-1\right],}$$

where X is ¹³C or ¹⁵N and R = ¹³C/¹²C for carbon and ¹⁵N/¹⁴N for nitrogen.

(c). Data analysis

We ran an analysis of similarities (ANOSIM; [29]) based on log-transformed fish density data (individual m⁻³) using a Bray–Curtis dissimilarity distance matrix to test whether tidepool fish assemblages differed between reef types. Additionally, the Hill numbers framework [30] was employed to compare the diversity of tidepool fish assemblages between reef types. The Hill numbers were calculated based on fish density (individual m⁻³) using ‘hillR’ package [31]. By altering the parameter q, where q = 0 corresponds to species richness (total number of species), q = 1 to Shannon entropy (Shannon diversity index) and q = 2 to inverse Simpson (Simpson’s diversity), a comprehensive spectrum of diversity measures is obtained considering the weighting given to rare and common species. A comparative analysis of the three indices (q0, q1 and q2) between reef types was conducted employing Brunner–Munzel tests [32] in the ‘brunnermunzel’ package [33].

The trophic structure of communities (i.e. fish and invertebrates) across both reef types was examined through the evaluation of the resource use, community niche partitioning, food web (trophic) hierarchy and trophic position of species. Resource use was evaluated by comparing the community δ¹³C values of consumers between reef types using the Mann–Whitney–Wilcoxon test. The community niche partitioning was examined through the proportional overlap between the modelled bivariate ellipses of δ¹³C and δ¹⁵N values of each community using ‘siber’ package [34]. An overlap > 0.6 was considered significant (indicating resource sharing), as defined by the classical overlap index [35] and frequently adopted in stable isotope studies [36,37]. The trophic hierarchy of consumers was defined based on their trophic level and position in the food web for the two reef types. The trophic position of consumers was estimated via the ‘tRophicPosition’ package [38], which incorporates a Bayesian model for multiple baseline (pelagic and benthic) estimates. In our study, benthic sources were macroalgae and pelagic sources were POM. The ‘tRophicPosition’ package (‘pairwiseComparisons’ function) was also used to investigate the potential differences in the trophic role of co-occurring species, that is, those species occurring both in biogenic and rocky tidepools. In this case, the probability that a species has posterior trophic position estimates higher, equal or lower between reef types was tested. All analyses were conducted within the R environment [39].

3. Results

Sixteen fish species (12 in biogenic and 10 in rocky tidepools) were recorded in the tidepools of the Santa Barbara Island (electronic supplementary material, table S1). This corresponds to 52% of the known tidepool fish diversity for the Abrolhos archipelago (i.e. 31 species). All but one (94%) were primary and secondary residents (electronic supplementary material, table S2).

Tidepool fish assemblages differed significantly between reef types (ANOSIM; p < 0.01) with Bathygobius geminatus markedly more abundant in biogenic tidepools and Tomicodon australis and the transient Mugil curvidens only occurring in rocky tidepools (electronic supplementary material, table S1). Brunner–Munzel tests revealed that there is no difference in ⍺-diversity between reef types (electronic supplementary material, table S3; figure S1).

Biogenic tidepool consumers showed more ¹³C-depleted values compared with rocky consumers (−12.01 versus −9.32 mean δ¹³C, respectively; electronic supplementary material, table S1; figure 2a). Resource use (δ¹³C values) differed significantly between consumers of both reef types (Mann–Whitney–Wilcoxon test; p < 0.01) and community niche overlap was low (<0.6), indicating niche diversification between reef types. The trophic hierarchy differed between biogenic and rocky tidepools with top consumers in the former being the green moray Gymnothorax funebris and in the latter the small-sized fillfrin goby Bathygobius soporator; in both cases, the basal consumers were amphipods (figure 2b). Only four species inhabited both biogenic and rocky tidepools (B. geminatus, Scartella cristata, Stegastes variabilis and Abudefduf saxatilis). Posterior estimates of trophic position values in both reef types indicated that species play different trophic roles in biogenic and rocky tidepools (figure 3). In all intraspecific comparisons, Bayesian posterior estimates showed that rocky tidepool specimens had greater than 90% probability of being at a higher trophic position than biogenic tidepool specimens.

Figure 2.

(a) Values of δ¹³C and δ¹⁵N in consumers of biogenic and rocky tidepools at Santa Barbara Island, Abrolhos archipelago. (b) Means (± s.d.) of trophic position estimates of biogenic (left) and rocky (right) tidepool consumers. Circles denote fish and triangles denote invertebrates. Amphipoda silhouette made by Hans Hillewaert under CC BY-SA 4.0. Fish silhouettes by R. Andrades.

Figure 3.

Comparative box plots of posterior estimates of trophic position between biogenic and rocky tidepool fish species. Fish silhouettes by R. Andrades.

4. Discussion

In the marine environment, small-scale spatial divergence in ecological patterns has been previously observed. However, the concept of small scale has varied substantially among authors, ranging from hundreds of metres to a few kilometres depending on the habitat under scrutiny. For instance, Le Bourg et al. [40] demonstrated that food chains differed between neighbouring coral reef habitats, while Wyatt et al. [41] and Skinner et al. [42] noted distinct trophic patterns across the reef seascape. To the best of our knowledge, no study had detected community-level changes in faunistic composition and trophic patterns of both slow- and fast-moving (i.e. fishes) consumers at a metre scale.

In the late 1940s, seminal works described the ecological variation occurring on the intertidal continuum of tropical limestone-rocky platforms, with an exclusive focus on emersed (non-pool) biota [9,22]. However, early scientific knowledge on intertidal rocky shores was mostly built based on studies in cold-temperate regions, as tropical shores have remained understudied to this day [8,11]. A few notable exceptions have yielded intriguing evidence that the composition of slow-moving and sessile biota inhabiting tropical intertidal reefs exhibits significant variation at the metre scale [43–45]. Here, we confirm this pattern for the intertidal community and provide evidence of taxonomic and trophic divergences for fast-moving organisms (fishes).

Closely set tidepool communities differed in their resource use, with the biogenic consumers possibly relying on heavily ¹³C-depleted pelagic nutrients and rocky consumers on less ¹³C-depleted benthic sources [41,46–48]. The biogenic intertidal reef is lower and closer to subtidal than the basaltic reef and this difference in height and distance to subtidal may sustain a higher access to pelagic subsidies. Notwithstanding, the importance of pelagic-driven resources needs to be further investigated to support this premise.

The divergence between trophic positions occupied by species occurring on both reef types sheds light on the ecological consequences for consumers inhabiting these two adjacent but distinct tidepool habitats. Trophic position is a quantitative trait that has evolved through natural selection and can be shaped in the context of food web composition and architecture [49,50]. The observed discordance between trophic positions may be the result of multiple factors given the niche diversification between the studied tidepool biotas. Variation in prey and food resource abundance, diversity or trophic structure between biogenic and rocky tidepools could explain the higher trophic positions of the rocky tidepool individuals [49]. For example, rocky tidepools in our study exhibit low algae coverage and amphipod density than biogenic ones (R Andrades 2025, unpublished data).

In fluctuating coastal environments, such as intertidal sandy and rocky shores, consumers adjust their dietary intake and trophic position in response to habitat variability [51]. For instance, fishes belonging to the genera Bathygobius, Scartella, Stegastes and Abudefduf exhibit more generalist (omnivorous) trophic strategies in tidepools [52–57] compared with their conspecifics in subtidal zones [58,59].

We have demonstrated that in tropical intertidal flat reefs, small-scale variation in community structure at the composition and trophic level can occur in response to subsidies to consumers. Stable isotopes have proven to be effective tracers for demonstrating trophic divergence in intertidal environments. In comparison with other dietary proxies, consumer isotopic composition provides a longer time-integrated measurement [60–62], further reinforcing our findings.

Ethics

This study was conducted in accordance with ethical guidelines and Brazilian legislation, including authorization to collect the studied organisms, under the permit no. 65951 (ICMBio/ MMA—Brazilian Ministry of Environment).

Data accessibility

Raw data (Excel spreadsheet) was uploaded as electronic supplementary material.

Supplementary material is available online [63].

Declaration of AI use

We have not used AI-assisted technologies in creating this article.

Authors’ contributions

R.A.: conceptualization, data curation, formal analysis, investigation, methodology, project administration, writing—original draft, writing—review and editing; H.C.G.: investigation, writing—review and editing; T.G.: resources, writing—review and editing; J.-C.J.: resources, writing—review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

No funding has been received for this article.