INVESTIGATOR: Scott M. Gallager Biology Department Woods Hole Oceanographic Institution Woods Hole, MA scott@videoking.whoi.eduData from this project undermines the present dogma that the primary prey of young cod larvae are copepod nauplii. Protozoans and other soft bodied microzooplankton are consumed in preference to copepod nauplii through at least the first 10 days of larval life. Cod larvae fed protozoans have significantly higher survival rates and utilize their yolk more sparingly compared with larvae fed nauplii alone. These data show protozoans to be essential in the larval diet if high survival is to be achieved. Numerical models of larval cod growth, survival, and distribution on Georges Bank must consider microzooplankton as potential prey.
Grazing by newly hatched cod larvae on microzooplankton prey. During the past two winter seasons (1993-1994, 1994-1995), a total of 24 experiments have been performed on newly hatched through 25 days post-hatch cod larvae. This required setting up a brood stock of 35 adult cod at the Environmental Systems Laboratory and manipulating environmental conditions to allow spawning at weekly intervals. Experiments consisted of short- term (3 h) grazing trials on three types of prey (dinoflagellates, protozoans, and copepod nauplii) with larvae subsampled from five long-term growth and rearing studies. Prey provided in the growth trials consisted of dinoflagellates, two groups of protozoans (tintinnids and oligotrichs), copepod nauplii, a mixed treatment consisting of all prey at equal biomass levels, and a starvation treatment. All prey were reared in the laboratory from mono-specific isolations from the field.
Since hard tissues are lacking in protozoans which would remain in the larval guts following ingestion, a new technique was developed for quantifying ingestion rates on soft-bodied prey. Prey were pre-stained with fluorescent cytoplasmic stains of three different colors then added to a series of larval containers at 6 increasing concentrations. After a period of 3 h (optimal time was predetermined which minimized egestion and digestion), larvae were sampled and gut fluorescence quantified using epifluorescent microscopy coupled with image processing techniques. To allow estimation of grazing rates, gut fluorescence was calibrated against the fluorescence of whole, stained protozoans at the beginning and end of each experiment.
Briefly the results are as follows: Newly hatched cod larvae (day 0) consumed soft bodied prey at a rate of about one to 10 prey per 3 h trial; no nauplii were ingested. Day 2 larvae, whose mouth had fully opened, consumed protozoans at a rate of 10 to 100 per 3 h trial. The number of protozoans consumed increased exponentially with time until day 15 when rates began to level off. Nauplii were first observed in the larval guts on day 5 but were not ingested at significant and reproducible rates until day 12. Thereafter until day 25, both nauplii and protozoans were ingested at significant levels. Complete morphomentrics and gut and mouth developmental stages were documented for each larva for which ingestion rates were measured.
Long-term growth and survival of cod larvae feeding on the various prey treatments gave some interesting clues as to the importance of protozoans in the diets of larval cod. There were few differences in growth rates (measured as increases in standard length, eye diameter, and myomere height) between the prey treatments. However, yolk sac volume decreased more slowly and survival was highest (>75%) in those treatments fed protozoans from hatching through 25 days.
3-Dimensional foraging behavior of larval cod feeding on microzooplakton prey. A number of numerical simulation models are being developed to forecast the distribution of gadid larvae in the field. Besides simulating larval growth, these models must also take into account energy balance based on the available prey field. Our 3-dimensional foraging studies are documenting the dynamics of the foraging process (swimming speed, attack speed, attack frequency, etc.) as a function of prey type and density which may be used in simulation models. 3-Dimensional optical imaging techniques are being used in the laboratory to quantify the foraging process of larval cod. My Post-Doc (Dr. Ione Hunt von Herbing) has spent two seasons making recordings of larval cod feeding on various prey at different densities and is in the process of working up the data. These data will be provided shortly to GLOBEC investigators (e.g., Dr. Francisco Werner) presently developing IBM models of larval cod distribution on Georges Bank.
Ship-board Experiments
Grazing. To evaluate prey selection by larval cod from natural assemblages of prey, we have been conducting short-term grazing experiments at sea. In addition, we have been documenting microzooplankton prey motility patterns from natural assemblages across Georges Bank. We conducted these experiments during the pilot cruise to Georges Bank in 1994 (CI9407) and the 5 GLOBEC Processes cruises in 1995 (EN259, EN262, EN264, EN266 and EN268). During these cruises, water was collected non-destructively from various locations in the water column. Cod eggs were spawned and fertilized in the laboratory one to two weeks ahead of the cruise and incubated between 2 and 10oC. Embryonic development was timed so that larvae would be hatching throughout the 15 day cruise period. Cod eggs were obtained from our broodstock at WHOI and from the St. Andrew's Fisheries Laboratory c/o Dr. Ed Tripple. Eggs were held on ship board in plastic containers and cleaned every two days. Excellent survival was obtained by keeping egg densities below 0.5/L and transferring to clean water regularly.
Two meter-long spar buoys equipped with lights and radar reflectors were used as drifters during the grazing experiments. Fifteen two liter polycarbonate bottles were arranged into three plastic milk crates and hung below the spar buoy on a two meter- long bungie cord and 3/4 nylon line. Three replicate bottles containing larval fish at a given age were set-up for each of the four treatments given below.
After the ship was positioned near the GPS drogue, the incubator was deployed through the A-frame by lowering the milk crates into the water first followed by the spar buoy. Most incubations began about 1000 hrs and were retrieved by 1700 hrs. Following incubation, larvae were removed from the bottles, mounted on slides and examined under epifluorescence microscopy using either blue or UV excitation for AO or Cell Tracker blue, respectively. Fluorescent images of larval guts were captured and stored digitally to allow quantification of gut fluorescence. Standard morphological measurements were also made on the stored image (length, height, yolk sac area, myotomal height, eye diameter, etc).
While the grazing experiments were underway, stained prey (both protozoans and nauplii) were held on the ship under conditions similar to those on the drifter incubator. These prey were used to calibrate the staining process and allow a specific- illuminance value to be assigned to individual prey. To obtain number of prey ingested by the larval cod, the fluorescence intensity at a specific wavelength (integrated illuminance values 0-255 for each pixel above a certain threshold) in the larval guts is divided by the specific-illuminance for a given prey item. A total of 26 deployments were made using these techniques.
Additional grazing experiments were conducted on shipboard to determine the effect of prey and predator (larval fish) density on grazing activity of the cod larvae. Three-day-old larvae were placed into the two L bottles at densities of 200, 100, 50, 25, 10 per bottle and allowed to graze for six hours on Balanion sp. (protozoan) at a concentration of 1 ml-1.
Results show that greatest grazing was seen at a larval density of 50/bottle, but only small decreases even at the extreme density of 200/bottle. Visual observations indicated a feeding frenzy occurred at larval densities greater than 10/bottle (i.e., feeding was encouraged by the high density of larval cod).
Results of the grazing experiments on various size fractions showed that newly hatched cod larvae feed directly on natural assemblages of microzooplankton, and particularly on protozoans. No copepod nauplii were ingested before day 6 following hatching. Feeding rates on Balanion sp. were comparable to the highest levels observed in laboratory experiments, but feeding rates on nauplii were lower than expected even in the enhanced treatments.
Prey Motility Experiments. The purpose of these experiments were to observe, record and analyze motility patterns and the size spectrum of available prey from two to three locations in the water column- near bottom, pycnocline, and upper well-mixed area on each of the 6 cruises. This was particularly important at the times when water samples were taken for the larval grazing experiments.
Water samples were collected from the near bottom, 20 m and 1 m below the surface with Go-Flo bottles. Samples were also collected from the surface with a bucket over the side. Go-Flo samples were either collected from the port as usual, or to the test the idea that microplankton are disrupted by this procedure, by siphoning from the bottle through the air port. 200 ml tissue culture flasks were filled with the sample and placed into an incubator at 5oC.
A B/W high-res Pulnix camera was fitted with a 50 mm macro lens and mounted on a frame across from a fiber optic ring illuminator fitted with a far-red filter. The entire apparatus was suspended within an incubator by bungie cord to reduce vibration produced by the ship. Recordings were made on SVHS medium for a period of 15-30 min for each sample. The flask was then replaced with the next sample and recordings continued. The field of view was set to 8 mm (scale bar at the beginning of each tape). Concurrently with the video recordings, the signal was sent to an image processor which processed images at about 1/sec for particle concentration, size, area, circularity, and a number of other morphological descriptors. A minimum of 200 data points were collected from each sample type.
Motility patterns are being analyzed with the Motion Analysis EV system in my lab at WHOI. The final output will be particle size distribution and a motility 'energy' spectra associated with each particle. This will be compared with species composition in the microzooplankton fraction preserved in Lugols and processed by Dr. Dian Gifford.
Data from both the laboratory and field experiments are being processed and tabulated. During the upcoming year, we plan to produce a series of manuscripts describing the following: