The Impact of Benthic Filter and Suspension Feeding Invertebrates on Water Column Phytoplankton Populations.

Topical Area: Algal Blooms, Zooplankton and Phytoplankton Ecology

Gabriel A. Vargo and Merrie Beth Neely, University of South Florida, Department of Marine Science, St. Petersburg, FL; Robert Erdman, Natalie Nickerson, and Keith Skorewicz, Eckerd College, St. Petersburg, FL; Gary Kleppel and Karen Roberts Kirtland, University of South Carolina, Columbia, SC; and Michael Savarese, Florida Gulf Coast University, Naples, FL.

Water column microalgal blooms of a small cyanobacterium, Synechococcus sp. were first noted in Florida Bay in 1987 following the die off of seagrasses (primarily Thalassia testudinum). Cyanobacterium blooms occurred in the central portions of the bay with diatom blooms in the western regions.

One hypothesis that has been suggested to account for the increased magnitude of the phytoplankton biomass is a reduction in loss rates due to reduced grazing by water column phytoplankton populations by zooplankton and by benthic filter and suspension feeders. A major loss of benthic filter feeding and suspension feeding organisms, particularly sponges, occurred in the central and southern portions of the Bay in 1991-1992. Such information led us to suggest the following hypothesis:

Reduced grazing activity by benthic filter and suspension feeders has contributed to the maintenance and prolongation of water column phytoplankton blooms.

A pilot study was undertaken to test this hypothesis at Keys Marine Laboratory during June, 1996 with a subsequent follow-up series of experiments in July, 1997. A final assessment will be made in July, 1998. The purpose of the study was to determine the magnitude of filtering and ingestion rates of representative benthic suspension and filter feeding invertebrates common to Florida Bay; to test experimental protocols for obtaining these rates and recommend modified procedures; to determine the general groups of phytoplankton that were ingested in field collected samples and in laboratory grazing studies; and ultimately to estimate the impact of filter and suspension feeders on water column phytoplankton biomass in Florida Bay.

Seven species of benthic invertebrates were used in our initial experiments: The bivalves Argopecten irradians,and Chione cancellata; the mangrove tunicate, Ecteinascidia turbanata; and the sponges, Chondrilla nucula, Cinachyra sp., Halichondria sp. and Haliclona sp.. Five additional species were tested in July, 1997: the red sponge, Tedania ignis, and the black sponge (identification pending), two tunicates encrusting on seagrass blades (white and black, identification pending), and a fringed oyster also growing attached to seagrass blades (Pinctada sp.).

Our objectives for the July, 1997 study included utilization of a benthic chamber to encapsulate specimens in the field and determine in situ filtering and ingestion rates and to again determine rates for representative specimens of each species under both static conditions and in a closed flow-through system in the laboratory. To determine in situ filtering rates, benthic chambers were constructed that encapsulated an animal attached to the substrate and its accompanying water column. A positive displacement diaphragm pump was used to circulate the enclosed water to the surface, through a fluorometer and back to the chamber. A valving system allowed samples to be collected during the pumping procedure and a portable computer was used as a data logger to record changes in in vivo fluorescence which were taken to be indicative of changes in chlorophylla concentration. Filtering and ingestion rates in the static and closed laboratory systems were determined as previously described. Briefly, representative specimens of each species were exposed natural phytoplankton populations at a range of chlorophyll concentrations derived from water collected from 2 locations; an area east of Whipray basin, which had chlorophyll concentrations of 5 ug/l, and from just offshore of the KML, with a chlorophyll concentration of 0.7 ug/l. Incubation times in the static experiments were 15 to 30 minutes while the enclosed systems were circulated for 30 minutes after a baseline was determined without an animal in the chamber. Calculations of filtering and ingestion rates are based on control corrected initial and final extracted chlorophyll determinations and all values are normalized to an individual animal and dry weight. Additional samples were taken for determination of selected carotenoid pigments in the water samples and in selected individual specimens taken from the natural environment during their collection and from the laboratory grazing experiments after incubation at the appropriate level of chlorophyll concentration was complete. Qualitative determinations of the types of phytoplankton ingested will be based on HPLC analysis of selected carotenoid pigments.


Average filtration and ingestion rates for all the species used in laboratory studies during the July, 1997 experiments were within the range of values determined during the previous summer using both experimental protocols. During our initial experiments, filtering and ingestion rates, irrespective of the water type and chlorophyll concentration, ranged from 100 to 850 ml/g dw/hr and 0.08 to 2.11 ug Chl/g dw/hr, respectively. However, the maximum values for ingestion rates for Chondrilla nucula, Chione and Argopecten increased to 7, 2, and approx. 12 ug Chl/g dw/hr, respectively. Filtration and ingestion rates under static and/or enclosed system experiments for the new species tested were:

Black tunicate: 46 ml/g dw/hr and 0.45 ug Chl/g dw/hr; white tunicate: 20 ml/g dw/hr and 0.2 ug Chl/g dw/hr; Tedania ignis: 295 ml/g dw/hr and 0.97 ug Chl/g dw/hr; black sponge: 720 ml/g dw/hr and 1.1 ug Chl/g dw/hr; Pinctata sp.: 470 ml/g dw/hr and 1 ug Chl/g dw/hr.

We were not able to obtain filtration and ingestion rates using the in situ chambers. Placement of the chambers over the organisms stirred up the sediments to such an extent that initial and final chlorophyll values were not solely representative of water column concentrations. Therefore, all trials mainly measured the settling rate of resuspended particles. We are modifying the chambers for this coming season with a flexible membrane across the bottom that will only allow the organism to be inserted and therefore reduce the amount of resuspended material trapped in the chamber. The modified chambers will be used to determine filtration and ingestion rates of the larger sponges such as Spheciospongia, Callyspongia, and Ircinia campana.

Water samples and, in several locations, animals were collected for HPLC determination of carotenoid pigments. Water samples from Whipray Basin contained fucoxanthin and zeaxanthin, indicative of diatom and cyanobacteria populations, with occasional traces of 19-hex, indicative of prymnesiophytes. All sponges, tunicates, and bivalves used in the static and closed system grazing experiments in which Whipray Basin water was used had high concentrations of both fucoxanthin and zeaxanthin which indicates that they were removing cells from the water. Some indication of selective feeding or partitioning of cells was indicated by the lack of fucoxanthin in the black tunicate whereas the white tunicate contained both carotenoids with fucoxanthin an order of magnitude greater than zeaxanthin. The results indicate that most of the filter feeding invertebrates used in our experiments do ingest the cyanobacterium.

The impact of grazing by benthic filter and suspension feeding animals on the water column cannot be calculated because there is still a lack of quantitative areal estimates of their abundance, biomass and distribution. Our initial assessment was based on the areal distribution for Chione cancellata as determined by Turney and Perkins (1972) can, however, be made based on some assumptions and a limited amount of data. Turney and Perkins (1972) report areal distribution values for Chione cancellata in several areas of Florida Bay. Values range from 3 to 24 animals/ft2 for the interior region of Florida Bay which translates to 32-258 animals/m2. With a maximum filtering rate of approximately 200 ml/animal/hr, Chione could filter 6.4 to 51.6 liters/m2/hr or .64% to 5.16% of a 1m3 water column (1000 liters) in 1 hour. If we assume particle removal occurs over a 24 hr period then the rates become 154 and 1238 liters/day or roughly 10 to >100% of our theoretical 1 m3 water column.

A similar calculation can be made using the filtration rates for all of the sponges used in our study. The average daily filtration rate was determined to be approximately 18 liters/animal/day. Stevely and Sweat (1995) measured a numerical abundance of 4832 total sponges per hectare in the Long Key area (their Table 4). Based on the average filtration rate and assuming a 1 m water column above the 1 hectare area (10,000 m2), the sponges could filter 88,873 liters/day or approximately 0.9% of the total water column. Although this value appears to be low, recall that our filtration rates are only based on small sponge species; rates for the larger specimens have not been determined in our experiments. However, Reiswig (1974) has calculated filtration rates for larger species that are 5 to 10 time higher than our maximum rates. If we add the minimum Chione areal filtration rate given above to the equation then the total for particle removal by benthic filter feeders is on the order of 16% of the 1 hectare, 1m deep water column.

Our calculation can be viewed as an underestimate because filtration rates of the larger sponges are not included in our total estimate. When losses from benthic suspension feeding are combined with water column grazing by zooplankton and microzooplankton, the total grazing loss could be represent a significant fraction of the overall phytoplankton daily productivity and growth. Consequently, loss of either or both of these trophic levels - as is suggested to have occurred during the peak of the Cyanobacterial blooms - would lead to a reduction in the loss rate of the phytoplankton community and subsequently a higher standing crop (biomass).