Projeto de pesquisa

To identify gaps and trends in scientific research in the GARS area, we will use an automated text analysis known as topic modeling. This method is often applied in social science but in recent years has become more used in biological sciences (Syed et al. 2018; Luiz et al. 2019; McCallen et al. 2019; Andrew et al. 2021). We will search articles found in Clarivate Analytics Web of Science using the following string (i.e., word combination): ((mesophotic OR shelf-edge) AND (reef OR ecosystem OR coral)) OR ("deep* reef"). The analysis will follow the methods from Luiz et al. (2019), using latent Dirichlet allocation (LDA) modeling to identify common topics in which the papers are grouped. Using the R package “ldatuning” (Murzintcev, 2014) we will estimate the most preferable number of topics for the model and define the number according to the optimization method proposed by Deveaud et al. (2014). To fit the best LDA model we will use the R package “topicmodels” (Hornik & Grün, 2011). Following the approach of Luiz et al. (2019), we will identify topics with increasing (hot topic) and decreasing (cold topic) popularity, topic specificity/generality (if they co-occur with other topics), and the research gaps concerning the mesophotic ecosystem and subsequently the GARS.
The functional traits of reef fish are widely studied (Luiz et al 2012; Luiz et al 2019; Mazzei et al 2021; McLean et al 2021), while the studies of benthic organisms traits are increasing in the last years (Lam-Gordillo et al 2020a, b). Here, we aim to assess the role of the AOP filter for the community, using a cross-taxa approach (Sunday et al. 2015; MacLean et al. 2017), encompassing species from the Caribbean and Brazil, including those from the GARS. To assess the role of the AOP as a filter, we will evaluate which traits allow organisms to cross the AOP, by analyzing the relationship between species traits and their distribution using generalized linear mixed-effects models (Luiz et al. 2012).
Adapting the methods from Tamir et al. (2019), we plan to do belt transects at stratified according to depth (every 10m), following the boat drift line, parallel to the shoreline. We will also film as detailed as possible the sites where the individuals of interest were registered. Footage acquired in previous cruises will be analyzed together with the videos acquired by this project. The data to be analyzed comprises videos from shallow areas (60m), taken using a double Deep Worker submarine and ROVs (see Francini-Filho et al. 2018).
L. purpureus will be used as a model organism to investigate the ontogenetic migration of species from the coast to the GARS. Individuals will be sampled across-shelf, targeting a wide range of length, covering juveniles and adults, to assess the ontogenetic shifts of habitat use, which is observed in other snappers (Aburto-Oropeza et al. 2009; Tanaka et al 2011). Additionally, we will also perform stable isotope analysis of otoliths to support the data on ontogenetic migration (Gillanders 2005; Wallace et al. 2014).
The data gathered by this Ph.D. project and other research in the region will contribute to an effective MPA network design (Mills et al. 2010), aiming to reconcile sustainable economic activities and biodiversity conservation. Using the R package prioritizr (Hanson et al., 2017; Magris et al. 2020), we will assemble data on species distribution, threats, and connectivity to support the definition of the PACs, covering the distribution of the species among the different habitats, allowing the species connectivity between them.