A Functional Perspective for Reef Benthic Communities: Temporal Trends in the Only Atoll in Southern Atlantic

Reef benthic communities provide many important ecosystem functions such as nutrient cycling, carbonate accretion and tridimensional complexity. Yet, reefs worldwide face an uncertain future, being threatened by local and global impacts. As an alternative approach to evaluate communities’ changes, functional ecology aims to understand how species shape the environment and how functions conduct ecosystems’ dynamics. The aim of this study was to investigate the temporal dynamics (2013–2019) of the reef benthic community in the most pristine reef in Brazil, Rocas Atoll, using a functional diversity approach. We identied 48 organisms grouped into 17 functional entities (according to their traits’ combination), considering all sampling years. Benthic community was temporally dominated by functional entities responsible for providing low reef tridimensional complexity (represented mainly by turf algae). This dominance reected in low values of functional entropy, due to uneven abundances distribution between unique functional entities, those that have unique trait combination. Functional richness oscillated over years, but did not show great changes in functional spaces, maintaining an equity in the number of functional entities and indicating stability of reef functions in Rocas Atoll, even with unequal abundances’ distribution. Our study is the rst to use a functional approach in temporal scale and represents a baseline for South Atlantic, since it provides the actual state of reef benthic communities using a functional approach, in an environment with no direct anthropic impacts. This can help to predict the effects on some ecosystem functions caused by local and global changes and its consequence for ecosystem services.


Introduction
Despite covering small areas in the oceans (Spalding and Grenfell 1997), reefs are the richest marine ecosystems, being comparable to tropical forests (Connel 1978;Costanza et al. 1997). In reefs, benthic communities deliver many important ecosystem functions (Solan et al. 2004;Brandl et al. 2019), such as nutrient cycling (Costanza et al. 1997), carbonate accretion and tridimensional complexity (Zawada et al. 2019a). Tridimensional complexity is a reef function that directly affects the habitat formation for many species, dictating how organisms interact with the environment and with each other (Zawada et al. 2019a). Yet, reefs worldwide face an uncertain future towards changes in the Anthropocene (Williams and Graham 2019), being threatened by local (e.g. over shing and pollution) and global (e.g. climate change) impacts (Bellwood et  As an alternative approach to evaluate communities' changes (Mouillot et al. 2013), functional ecology aims to understand how species shape the environment and how functions conduct ecosystems' dynamics . By using species traits, functional diversity has become increasingly appropriate to evaluate and to monitor communities (Ernst et al. 2006), by accessing spatiotemporal changes in different community's aspects through functional diversity indices (Mouillot et al. 2013;Biggs et al. 2020). On marine environments, functional approaches have been widely applied for shes and invertebrates (Bremner et al. 2003(Bremner et al. , 2006 Benthic communities of Brazilian reefs are dominated by turf algae and macroalgae (Aued et al. 2018). These reefs are threatened by many disturbances, as those associated to urban development, coastal runoff, marine tourism and overexploitation of reef organisms, and by ocean warming effects (Leão et al. 2019). Even in oceanic islands, which are relatively isolated from anthropogenic effects, it is possible to observe global scale impacts (Ferreira et al. 2013). For being unique and remote systems, oceanic islands reefs may operate as natural laboratories and as a reference for monitoring how these ecosystems behave through distinct spatiotemporal scales Gatti et al. 2015). Moreover, spatiotemporal studies and monitoring programs support the identi cation of long-term changes in reef benthic communities as seen in the Caribbean, for example (Gardner et Knowlton and Jackson 2008). In this context, a long-term ecological research can help to ful ll gaps of knowledge in understanding reef communities' spatiotemporal patterns in the Brazilian oceanic islands. All four Brazilian oceanic island are Marine Protected Areas, nevertheless, some are already impacted by over shing, tourism (Giglio et al. 2018) and by coral bleaching events associated to thermal anomalies (Ferreira et al. 2013;Gaspar et al. 2021). These phenomena jeopardize reef biodiversity, and ecosystem functions and services (Hughes et al. 2018a; Williams and Graham 2019). However, an understanding of temporal patterns in ecosystem functions of oceanic islands is still lacking. Here, we used the functional approach to investigate the temporal dynamics (2013-2019) of the reef benthic community in the most pristine reef in Brazil, Rocas Atoll. To date, this is the longest publicly available dataset for benthic communities in the Southwestern Atlantic, representing a baseline for future studies of dynamics and functional diversity from reef benthic communities.

Study area
Rocas Atoll (03°50'S, 33°49'W) is located approximately 267 km off the Brazilian coast (Kikuchi and Leão 1997). It is considered one of the world's smallest atolls, with 3,7 km in E-W directions and 2,5 km in N-S (Soares et al. 2011), and the only atoll in the South Atlantic Ocean (Kikuchi and Leão, 1997). Crustose coralline algae (CCA), vermetid gastropods and foraminiferans are the main shallow reef builders, while corals, mostly represented by Siderastrea stellata, have lower coverage and contribute less to the reef framework structure (Gherardi 1995;Kikuchi and Leão 1997). Rocas Atoll is the rst Brazilian no-take marine reserve, established in 1979. It has controlled anthropic impacts (e.g. very low pollution and shing) and it is probably the most similar to a pristine reef in Brazil, being a natural laboratory and an important study area (Longo et al. 2015).

Benthic community sampling
Reef benthic community was annually monitored from 2013 to 2019 (May and June). In each year, benthic data was collected by photoquadrat method in four different sites: Âncoras, Cemitério, Falsa Barreta and Podes Crer (Fig. 1). Every year, three 20 m transect were xed at each site, and a 1 m 2 quadrat was photographed every 2 m along each transect. In summary, over 838 photoquadrats were taken from the same reef area across seven years ( Table 1 in Appendix). Percent cover of each photoquadrat were analyzed with CPCe software (Kohler and Gill 2006), randomly distributing 50 points over each photo. The organisms below each point were identi ed to the lowest taxonomic level possible.  Functional diversity indices and data analysis  (Table 3).  Functional Entropy (FEnt) was low every year, indicating uneven percent cover between functional entities (Fig. 6). Despite signi cant oscillations in FEnt between years (GLM p < 0.05), differences among years occurred in small scales (Fig. 6), showing small variations in percent cover of functional entities with unique trait combination. These uctuations in FEnt were conducted by low and high tridimensional complexity and encrusting bioconstructors functional entities percent cover, located at functional space vertices. The highest FEnt was in 2013, representing a well-distributed percent cover of extreme functional entities if compared to other years, while the lowest was in 2019, due to uneven percent cover distribution, when reef benthic community was dominated by low tridimensional complexity functional entities represented mainly by turf algae (59.19% cover; Fig. 3).

Discussion
Our study is the rst to use functional approach in a temporal scale for reef benthic communities. Our results show that Rocas Atoll benthic community is temporally dominated by functional entities responsible for providing low reef tridimensional complexity (e.g. turf algae, cyanobacteria and Zoanthus sociatus), high tridimensional complexity and primary production (e.g. macroalgae as Sargassum spp., Dictyopteris spp.), with low contribution of carbonate accretion (e.g. crustose coralline algae and corals as Siderastrea stellata, Porites astreoides and Favia gravida) and nutrient cycling (e.g. sponges and/or ascidians) functional entities. This dominance re ected in low values of functional entropy (FEnt), due to uneven abundances distribution between unique functional entities, that is, those that have unique trait with intense water circulation and wave action that can in uence the availability and sediment dynamics (Hearn et al. 2001). For example, depending on the quantity of sediment, it can cause shading and hence, affect primary production (Riul et al. 2008), also it can promote the temporal burrowing and smothering of coral colonies (Brown et al. 2002). This probably explains differences in percent cover of massive bioconstructors functional entities. The intense hydrodynamic at shallow tide pools can also favor the presence of more resistant functional entities (Steneck and Dethier 1994;Cheroske et al. 2000), such as the ones composed by epilithic algal matrix, macroalgae and crustose coralline algae.

Con icts of interest
The authors declare that they have no con ict of interest.

Availability of data and material
All data generated and analyzed during this study are available from the corresponding author. The LTER program will publish all data on a global database soon.

Code availability
The code used for the analysis are available from the corresponding author.
Author's contributions VAPP, AWA and BS: substantial contribution to the concept and design of the study; contribution to data collection; contribution to data analysis and interpretation; contribution to the manuscript's preparation.