Marine ecoregions and subecoregions within Indo-West Australian waters: A statistical approach based on species distributions



Aim: The Marine Ecoregions of the World (MEOW) system delineates the oceans into 232 ecoregions. Here, we aimed to evaluate the suitability of this system to represent species distributions within Indo-West Australian waters, explore alternative ecoregions and new subecoregions, and investigate environmental variables that are correlated with species distributions within those waters. Location: Indo-West Australia Taxa: Vertebrates, invertebrates, marine plants Methods: We downloaded occurrence data for 14,513 marine species from the Ocean Biogeographic Information System. We analysed differences in species composition among nine ecoregions within Indo-West Australian waters using pairwise permutational multivariate analysis of variance to evaluate how well the MEOW system represents species distributions within those waters. We delineated subecoregions within each distinct ecoregion using hierarchical cluster analysis with the unweighted pair-group method using arithmetic averages. We analysed relationships between environmental variables and species composition using distance-based linear models. Results: Species composition was significantly different among ecoregions, except for three adjacent regions, which were combined into a single large ecoregion. Hence, seven distinct ecoregions were further analysed. Our study identified 13 subecoregions within these ecoregions that each separate into ‘inshore’ and ‘offshore’ zones. Depth explained the most variation in species composition of the combined taxa and sea surface temperature was the most important parameter in explaining the variability in most taxa. Main conclusion: The MEOW system did not represent well the distribution of marine species within Indo-West Australian waters. Alternatively, we show that those waters encompass seven distinct ecoregions with 13 subecoregions. The main environmental drivers of species distributions could be depth and sea surface temperature. The proposed ecoregions and subecoregions allow us to improve the biogeographic hypotheses for understanding the evolution of marine species and identify representative marine habitats and species composition for the setting of Marine Protected Area networks within Indo-West Australian waters.,Study area The study area includes nine currently recognised ecoregions within Indo-West Australian waters: Western Sumatra, Southern Java, Lesser Sunda, Bonaparte Coast, Exmouth to Broome, Ningaloo, Shark Bay, Houtman, and Leeuwin. A map of these ecoregions is available from The study area was divided into a 1° grid-cell resolution to create 473 cells, including 83 grid cells in Western Sumatra, 70 in Southern Java, 71 in Lesser Sunda, 46 in Bonaparte Coast, 67 in Exmouth to Broome, 15 in Ningaloo, 25 in Shark Bay, 24 in Houtman, and 72 in Leeuwin. Species occurrence data Species of marine Mammalia, Reptilia, Pisces, Mollusca, Crustacea, Echinodermata, Annelida, Cnidaria, Bryozoa, Porifera, Alismatales, Chlorophyta, Rhodophyta, and Phaeophyceae were selected. Species data from Indonesian and Australian waters were downloaded from the Ocean Biogeographic Information System (OBIS) ( on October 20, 2019. Data were retrieved for 28,923 species and 1,497,347 location records. A total of 14,556 species from 81,968 cell records were extracted after overlaying species data and grid cells. Duplicate species within grid cells were removed. OBIS species names were matched with those in the World Register of Marine Species (, and names marked as synonyms in WoRMS were linked to currently accepted names. This data evaluation reduced the number of species from 14,556 to 14,513. Grid cells that contained fewer than five species were excluded to minimise distortion in similarity analyses (Kreft & Jetz, 2010). This reduced the number of grid cells from 473 to 338. The final grid cell counts by ecoregion were 54 in Western Sumatra, 33 in Southern Java, 43 in Lesser Sunda, 42 in Bonaparte Coast, 61 in Exmouth to Broome, 9 in Ningaloo, 25 in Shark Bay, 23 in Houtman, and 48 in Leeuwin. The same approach at both finer (0.5°) and coarser (2°) resolution was conducted as a supplementary analysis. In these analyses, the number of final grid cells was 816 (49.60% of total cells) and 135 (93.75% of total cells), respectively. Environmental data Environmental variables included both surface and benthic environments: salinity, nitrate, dissolved oxygen, current velocity, temperature, pH, photosynthetically available radiation (PAR), and depth. Salinity, nitrate, dissolved oxygen, current velocity, temperature, pH, and PAR data were downloaded from Bio-ORACLE (Tyberghein et al., 2012; Assis et al., 2018), and depth data were downloaded from Global Marine Environment Datasets (IOC et al., 2003). The mean values of environmental variables within grid cells were calculated and used in the analysis.,The ZIP file contains three datasets (ecoregion data_1.Rdata, ecoregion data_05.Rdata, and ecoregion data_2.Rdata) and README.txt. The README file contains an explanation of datasets.,
Date made available1 Jan 2021

Cite this