Interocean patterns in shallow water sponge assemblage structure and function

Sponges are a major component of benthic ecosystems across the world and fulfil a number of important functional roles. However, despite their importance, there have been few attempts to compare sponge assemblage structure and ecological functions across large spatial scales. In this review, we examine commonalities and differences between shallow water (<100 m) sponges at bioregional (15 bioregions) and macroregional (tropical, Mediterranean, temperate, and polar) scales, to provide a more comprehensive understanding of sponge ecology. Patterns of sponge abundance (based on density and area occupied) were highly variable, with an average benthic cover between ~1 and 30%. Sponges were generally found to occupy more space (percentage cover) in the Mediterranean and polar macroregions, compared to temperate and tropical macroregions, although sponge densities (sponges m^-2 ) were highest in temperate bioregions. Mean species richness standardised by sampling area was similar across all bioregions, except for a few locations that supported very high small-scale biodiversity concentrations. Encrusting growth forms were generally the dominant sponge morphology, with the exception of the Tropical West Atlantic, where upright forms dominated. Annelids and Arthropods were the most commonly reported macrofauna associated with sponges across bioregions. With respect to reproduction, there were no patterns in gametic development (hermaphroditism versus gonochorism), although temperate, tropical, and polar macroregions had an increasingly higher percentage of viviparous species, respectively, with viviparity being the sole gamete development mechanism reported for polar sponges to date. Seasonal reproductive timing was the most common in all bioregions, but continuous timing was more common in the Mediterranean and tropical bioregions compared to polar and temperate bioregions. We found little variation across bioregions in larval size, and the dominant larval type across the globe was parenchymella. No pattens among bioregions were found in the limited information available for standardised respiration and pumping rates. Many organisms were found to predate sponges, with the abundance of sponge predators being higher in tropical systems. While there is some evidence to support a higher overall proportion of phototrophic species in the Tropical Austalian bioregion compared to the Western Atlantic, both also have large numbers of heterotrophic species. Sponges are important spatial competitors across all bioregions, most commonly being reported to interact with anthozoans and algae. Even though the available information was limited for many bioregions, our analyses demonstrate some differences in sponge traits and functions among bioregions, and among macroregions. However, we also identified similarities in sponge assemblage structure and function at global scales, likely reflecting a combination of regional- and local-scale biological and physical processes affecting sponge assemblages, along with common ancestry. Finally, we used our analyses to highlight geographic bias in past sponge research, and identify gaps in our understanding of sponge ecology globally. By so doing, we identified key areas for future research on sponge ecology. We hope that our study will help sponge researchers to consider bioregion-specific features of sponge assemblages and key sponge-mediated ecological processes from a global perspective.