Vol. 519: 265–283, 2015 MARINE ECOLOGY PROGRESS SERIES Published January 20 doi: 10.3354/meps11093 Mar Ecol Prog Ser FREEREE ACCESSCCESS REVIEW A review of evidence for food limitation of sponges on Caribbean reefs Joseph R. Pawlik1,*, Steven E. McMurray1, Patrick Erwin1, Sven Zea2 1Center for Marine Science and Department of Biology and Marine Biology, University of North Carolina Wilmington, Wilmington, NC 28409, USA 2Universidad Nacional de Colombia, Centro de Estudios en Ciencias del Mar-CECIMAR, Santa Marta, 470006, Colombia ABSTRACT: The diversity and biomass of sponges on Caribbean reefs exceeds that of all other benthic organisms. Wilkinson & Cheshire (1990; Mar Ecol Prog Ser 67:285−294) summarized evi- dence that sponge communities on Caribbean reefs were different from those on the Great Barrier Reef, Australia, reflecting food limitation of the latter, but not of the former. Since then, we have learned much about the ecology of sponges, and the concept of food limitation has recently re- emerged, seemingly without substantive reference to past inter-oceanic comparisons or pioneer- ing studies of sponge physiology. Here, we review the literature in an attempt to correct this digression. Based on current knowledge of sponge nutrition and bottom-up resource gradients (particulate food, dissolved organic carbon [DOC], light, turbulence), we predict 4 patterns of sponge abundance that would suggest food limitation on Caribbean reefs. After a critical review of survey data, correlative studies and manipulative experiments, we can find no evidence for food limitation. Although there is good evidence for higher availability of picoplankton at greater depths, sponge abundance does not mirror this gradient, suggesting that other sources of nutrition are also important, and particulate food is not a limiting factor. Recent studies have renewed inter- est in the uptake of DOC by both high and low microbial abundance sponge species, suggesting that the absence of bottom-up effects for sponges on Caribbean reefs may be attributable to this mysterious, and likely ubiquitous, food source. The recent unambiguous demonstration of top- down effects of predation on sponge community composition across the Caribbean may owe its clarity to the relative lack of confounding influences of abiotic and bottom-up effects in this study system. KEY WORDS: Bottom-up · Top-down · Food webs · Suspension feeding · Symbiosis · Photoautotrophy · DOC · Benthic–pelagic coupling · Coral reefs · Sponge loop Resale or republication not permitted without written consent of the publisher INTRODUCTION munities of organisms are structured by some combi- nation of factors that are top-down (controlled by Ecologists face the daunting task of attempting to predation from the highest trophic levels) or bottom- explain the distributions and abundances of organ- up (controlled by the availability of resources such as isms that are influenced by a diversity of biotic and nutrients or food to the lowest trophic levels). The rel- abiotic factors. A fundamental theory used to address ative importance of these 2 factors has been vigor- this topic is food web dynamics, considered central to ously debated by ecologists for as long as the concept ecology (Fretwell 1987), which proposes that com- of food webs has existed (examples in Hunter & Price *Corresponding author: [email protected] © Inter-Research 2015 · www.int-res.com 266 Mar Ecol Prog Ser 519: 265–283, 2015 1992, Power 1992, Heck & Valentine 2007, Turking- phylogenetically diverse, morphologically variable, ton 2009). Marine ecologists have provided particu- and functionally clonal. The bottom-up resource of larly good characterizations of food webs with stud- importance for Caribbean sponges and the focus of ies of, for example, temperate rocky-shore intertidal this review is food, which can take the form of partic- communities (Paine 1969, Lubchenco 1978, Sousa ulate matter (including picoplankton), DOC, or pho- 1979) and tropical benthic macroalgal communities tosynthate from endosymbiotic microorganisms. At (Hay 1981, Lewis 1985). Due to the complexity of issue is whether limitation of the quantity of food most ecosystems, however, not only is some combi- available to sponges on Caribbean reefs is sufficient nation of top-down and bottom-up processes likely to to influence the overall biomass (a better term than be important (Hunter & Price 1992, Menge 2000, Ain- ‘population’ for a functionally clonal animal) of indi- ley & Hyrenbach 2010, Denyer et al. 2010), but also vidual species and the structure of sponge communi- the relative importance of these processes may change ties. Evidence for food limitation may be found in along with other factors, such as latitude (Marczak et direct measurements of sponge growth, or indirectly al. 2011) or time (Whalen et al. 2013). as correlative measurements of biomass, usually as In this review, we specifically consider whether the percentage cover of the reef substratum. Percentage community of sponges on Caribbean reefs is influ- cover is more useful than abundance (number of enced by bottom-up factors. This assessment is war- individuals), because of the wide range of sizes ex - ranted for several reasons. First, the importance of hibited by many sponge species. These measure- sponges on coral reefs worldwide is attracting more ments may be specific for individual species, or they attention as the relative abundance of reef-building may include all species in the sponge community. corals has declined (Bell et al. 2013). On Caribbean Changes in the sponge community can include an reefs in particular, sponge communities have greater overall increase in the biomass of sponges, or in the diversity and biomass than any other benthic organ- structure of the sponge community, as some species isms (Diaz & Rützler 2001) and sponge abundance is replace other species independent of changes in increasing (McMurray et al. 2010, Colvard & Ed - overall sponge biomass. Note that we will avoid the munds 2011, Ruzicka et al. 2013, Villamizar et al. use of the term regulation in this review because it 2014). Second, the topic of bottom-up control is made implies density-dependent control of a population, more complex because of the mechanisms by which and it seems unlikely that sponges are near their car- Caribbean sponges gain nutrition: in addition to sus- rying capacity on Caribbean reefs, either as individ- pension feeding on particles that include picophyto- ual species or as a community. It is also important to plankton and bacteria (Reiswig 1974, Trussell et al. recognize that, other than the resource of food, Car- 2006), sponges may also absorb dissolved organic ibbean reef sponges may be limited by space, as most carbon (DOC) (Reiswig 1981, Maldonado et al. 2012), are required to be attached to the substratum. And or may rely to some extent on photosymbionts (Erwin while we will consider light and water movement as & Thacker 2008, Freeman et al. 2013). Most recently, resource-related factors, we will not discuss other it has been proposed that sponges not only rapidly abiotic factors (salinity, temperature, sedimentation, absorb DOC from the water column, but also lose oxygen content, etc.) because variation in these much of the acquired biomass through the produc- parameters is generally low at the depths (>5 m) of tion of cellular detritus that is returned to the reef the area (slope of the fore-reef) where most reef ecosystem as part of a ‘sponge loop’ (de Goeij et al. sponges are found, but may strongly affect sponge 2013). Third, as has occurred with other study sys- populations in shallow water mangrove and grass- tems (Power 1992), a debate has emerged about the bed environments (Pawlik et al. 2007a, Nagelkerken relative importance of bottom-up and top-down fac- et al. 2008). tors in structuring the ecology of sponges on Carib- bean reefs (Lesser 2006, Trussell et al. 2006, Lesser & Slattery 2013, Pawlik et al. 2013). Here we will focus WHAT DO SPONGES ON CARIBBEAN REEFS EAT? specifically on bottom-up control of sponges on Ca - rib bean coral reefs, considering the evidence from The species diversity of sponges on Caribbean observational, correlative and manipulative studies. reefs is very high, with more than 100 conspicuous Before proceeding, it is important to briefly define species on most reefs within human access by scuba terms that are commonly used in discussions of food diving, and hundreds more species restricted to cryp- webs and trophic interactions as we will apply them tic locations or at greater depths (Diaz & Rützler to sponges, keeping in mind that these animals are 2001). Among these species are representatives of 3 Pawlik et al.: Food limitation of Caribbean sponges 267 poriferan classes, the Calcarea, Homoscleromorpha in situ have been developed, which, when coupled and Demospongiae, with over 12 orders in the last of with flow cytometry, have revealed that sponges dif- these classes. It is generally understood that sponges ferentially retain some microbial cell types over oth- are suspension-feeding organisms that use a combi- ers (Yahel et al. 2005). The use of stable isotopes and nation of their aquiferous system and the microvilli fatty acid signatures demonstrated that different on their collar cells to remove particles >0.1 µm from sponges within a community in McMurdo Sound, sea water (Riisgård & Larsen 2010), but that is not the Antarctica fed selectively on live cells of different full extent of their ability to feed themselves (Mal- classes