U.S. GLOBEC: Laboratory and Field Studies of Impacts of Invertebrate Predators on Fish Larvae on Georges Bank

Barbara K. Sullivan and Grace Klein-MacPhee

1. Goal

To determine the magnitude of predatory and competitive impacts of the hydroid Clytia gracilis on larval cod and haddock on Georges Bank.

2. Rationale

Research during Phase I identified hydroids as one of the invertebrate predators with the most potential to compete with and prey on larval cod and haddock on Georges Bank. This potential is strongly controlled by 1) degree of spatial and temporal overlap between predator and prey species , 2) abundance of the predators in the region of overlap, and 3) individual predation rates. This research will provide new information on each of these factors that will allow us to calculate predation induced mortality on larval cod and haddock.

3. Research Tasks

a) Determine patterns of abundance and distribution of hydroids in the larval fish patch. We will enumerate hydroids from bongo grid samples and selected MOCNESS tows to determine abundance and distribution patterns during April and May 1997 on the southern flank of Georges Bank. Samples will be collected by R. Greg Lough during his efforts to locate larval fish ( Lough et al., U. S. GLOBEC: Dispersive and Advective Influences on the Survival of Cod and Haddock Larvae on Georges Bank.).

b) Determine hydrographic influences on abundance and distribution of hydroids. Preliminary analysis of plankton samples from bongo grid surveys and MOCNESS tows from May 1995 indicated a high degree of spatial overlap with larval fish but high variability on both vertical and horizontal scales. Abundance varied several orders of magnitude with depth; abundance of hydroids was sometimes uniform throughout the water column but at other times was high only at depth. Our laboratory experiments have shown that hydroids do not swim and are negatively buoyant, sinking a rates of 0.3 - 3mmsec-1. Thus, hydroids must be kept suspended in the water column by tidal or storm induced turbulence. Indeed, they appear to be confined to well-mixed waters on Georges Bank. Highest numbers appeared to accumulate at the shelf-slope front on the southern flank of Georges Bank. Thus, hydrography is likely to be a key factor determining the degree of spatial overlap between hydroids and their prey. Our ability to predict predation induced mortality rates depends upon a clearer understanding of how hydrography controls distribution and abundance of hydroids, and thus, encounter rates between predators and prey.

We will use information on water column structure (from the CTD mounted on the bongo frame, ship mounted temperature and conductivity sensors, temperature data from subsurface drifters), current flow from ADCP and subsurface drifters to infer mechanisms controlling local vertical and horizontal distribution of hydroids . Maps indicating predator abundance versus longitude, latitude and bottom contours will be constructed. Eventually, information on current structure from fixed moorings will also be available to provide insight into physical processes that control distribution of these predators in the plankton.

c) Determine predation rates of Clytia on target species. During Phase I we showed that hydroids prey on copepods from the very definitive evidence of gut contents. We can calculate in situ feeding rates on copepods now (by using gut residence time together with frequency of prey in the gut). However, while we have evidence for predation on larval fish from laboratory experiments in small containers, we have no estimates of in situ predation rates on larval cod and haddock. Fish larvae are digested external to the guts and leave no identifiable remains. Experiments in containers smaller that 3 m3 are known to over-estimate predation rates on larval fish. Thus, we will measure predation rates in 13 m3 mesocosms to minimize container effects. Experiments will investigate the effect of prey size, age and species on predation rates.

d) Determine effects of turbulence and turbulent scales on predation rates. Measurements will be made to resolve turbulent velocities in the mesocosms. Measurement methods being used are outlined in Figure 1. We will use these measurements to direct optimal design of paddles to reproduce turbulent dissipation rates in the field. Because of the size of the mesocosms (1.8 m diameter, 5 m deep) we can generate turbulence at eddy scales of up to 1 meter. Eddies on this scale will distribute hydroids throughout the water column thus bringing them in contact with larval fish prey. [Eddy velocities at scales of 10 cm or greater should be 2-5 mm sec-1 at turbulent dissipation rates typical of Georges Bank, and are sufficient to overcome sinking rates of hydroids of 0.3-3mm sec-1. {D. Hebert personal communication}]

Smaller scale eddies, on the order of millimeters to centimeters, have been hypothesized to effect predation rates of planktonic predators either positively, through enhanced encounter rates, or negatively, through disturbances of the feeding process. Alteration of the design of mixing paddles could produce turbulence at scales less than 1 m; however, there is sufficient separation between meter and cm scales to expect that turbulent energy will cascade from the larger eddies down to a fully developed turbulent spectrum. There is good evidence for this presented in Fig. 1 which shows the expected - 5/3 slope of turbulent energy dissipation. Within the limitation imposed by the budget we will examine the influence of different scales of turbulence as well as different dissipation rates on predation rates.

e) Calculate hydroid related mortality of larval cod and haddock in the larval fish patch. Information on hydroid abundance and distribution, larval fish abundance and distribution, and feeding rates from the mesocosms will be used to calculate predation impacts in the region of the Bank sampled during April and May 1997.