Predation Group

Workshop Report

Members of the Predation Group attending the September 1998 SI Data Workshop were Steve Bollens, Larry Madin and Barbara Sullivan. Heidi Franklin from the University of Maine (working part time with Bollens) joined on Saturday to work with us on Matlab programming. The following topics were covered.

1. Review of progress on analysis of samples.

All of the Broadscale Survey MOC-10 samples from 1995 and 1996, and a small proportion of those fom 1997 and 1998, have been analyzed in BollensÕ lab, including identification and enumeration of all specimens, and measurement for length of up to 50 individuals of each species per net. Work will continue in BollensÕ lab on remaining 1997 and 1998 samples, and 1999 samples from Phase III. Representative plots of seasonal abundance for some larger organisms are attached to this report.

The number of MOC-10 samples from the Process cruises, 1994-1997 is shown below. All samples are given rough counts immediately after collection and before preservation. Lab counts have been completed on a subset of the preserved samples that represent replicated sampling in regions of interest. Remaining samples will be completed in 1998 and 1999, according to priorities established by the Predation Group. (See Figure A and Figure B.)

Number of MOC-10 samples collected:

Year Cruise # of tows
1994 AL9403 21
1995 EN268 22
SJ9503 6
SJ9505 5
SJ9507 12
1997 EN298 24
EN301 23

MOC-1 samples from process cruises in 1994, 1995 and 1997 have been analyzed in Sullivan's lab for identification and enumeration of invertebrate predators. Predators >10 mm in length are counted from the whole sample; smaller, more numerous predators are counted from subsamples. Predators are placed in size categories: 1-3 mm, 3-10mm, >10mm. Bongo net samples from April and May process cruises collected by R.G. Lough 1993-1997 have been counted for hydroids. Analysis of 1994 and 1995 MOC-1 samples is completed. Analysis of 1997 samples is in progress and has been done in cooperation with K. Wishner. A representative plot of vertical distribution of a small but ubiquitous and abundant predator of copepods collected during 1997 is available as Figure C.

Number of MOC-1 and bongo net samples processed:

Year MOC-1 Bongo
1993 NA 47
1994 16 62
1995 148 96
1997 91 66

2. Review of Predator List

We created a revised list of predators on Georges Bank, based on data from MOC-1 and MOC-10 collections. A comprehensive list of about 40 species will be narrowed to a short list of the most abundant predators, which we expect to number 10-15 species. The status of information on feeding rates and selectivity, including functional responses and temperature effects was also reviewed for each species. We have our own data on feeding rates or selection for about 10 species, and expect to use values from the literature for about another 10-15. Additional data on predation rates by copepods and other small predators will be obtained on Process cruises in 1999.


Aequorea sp.
Aglantha digitale
Clytia gracilis (hydroids)
Cyanea capillata

hydromedusae (misc)
Mitrocomella sp.
Nanomia cara
Pelagia noctiluca
siphonophores (misc)

Bolinopsis infundibulum
Pleurobrachia pileus

Sagitta elegans

Candacia spp.
Centropages spp.
Euchaeta spp.
Metridia spp.
Tortanus spp.

Meganyctiphanes norvegica
Nematoscelis sp.
Nematobrachion sp.
Thysanoessa inermis
other euphausiids

Neomysis americana

Crangon septemspinosa
Dichelopandalus leptocerus
Pasiphaea multidentata

pagurid megalopae
Sergestes arcticus

Byblis serrata
Gammarus sp.
Monoculoides edwardsi
Themisto gaudichaudii

Cirolana polita

Ammodytes sp.
Clupea spp.
Cyclothone spp.
snake mackeral

3. Matrix for calculation and plotting of predator and prey distributions, and predation related mortality.

We discussed the structure of a general matrix for calculating predation mortality and mapping it in time and space, based on distributional data for predators and prey, and appropriate feeding rate information. Distribution data will come mainly from the Broadscale MOC-1 and MOC-10 collections, supplemented in some places with data from Process cruises (MOCNESS or bongo tows). Rate data will be from a combination of our own measurements, from feeding experiments or gut contents analysis, and comparable results from the literature for other species. All rate data will be normalized to an arbitrary temperature (i.e. 10°C), and then adjusted to ambient temperatures for different seasons and locations. The calculation matrix includes:

The expected output of the matrix is an estimated predation mortality for each predator-prey pair, at each Broadscale station. The matrix should allow depth-specific calculations if desired. It will then be possible to sum the effects of all predators on a given target species, to map the total predation mortality for that species in space and time on Georges Bank. These values can be compared to demographic estimates of mortality, and incorporated into models of population dynamics.

During the Workshop we worked on programming the structure of the matrix in Matlab. Initial runs can be tried in the near future. Data will be mapped using the Matlab contouring and mapping routines under development in the Georges Bank program.

4. Gut Contents Analysis

We still need to complete a large number of microscopic gut-content analyes for the major predators. While we have considerable data for Crangon and Pleurobrachia, Sagitta and hydroids, and some for Themisto and Bolinopsis, there are many samples of other predators remaining to be done. Recent examinations by Sullivan of small Themisto, for example, showed that copepod remains were identifiable in the guts. Existing and additional data on gut contents will be used with our own and other rates of gut passage to estimate in-situ feeding rates.

5. Antibody and DNA probes for gut contents.

The antibody reagents we have developed over the last 4 years have recently been shown to be less specific to Calanus than earlier thought. Specificity appears to depend on carefully controlled dilutions and other conditions, which cannot always be met with field-collected samples. We plan now to use the antibody, which is specific to copepods, to detect the presence of copepods in the guts of predators, such as euphausiids, mysids and gammarid amphipods, whose diet we know little about. This will enable us to limit our list of predators on target species.

Support for use of DNA probes to identify cod, haddock, Calanus and Pseudocalanus remains in predator guts was cut from our Phase III proposal. However, Madin has a new student, Annette Frese, who has experience with these techniques, and is interested in applying DNA probes to the analysis of gut contents. We anticipate cooperation with Ann Bucklin in this work, and will probably be obtaining samples during the 1999 Process cruises.

6. Use of Process cruise samples.

Sampling with the MOC-1 and MOC-10 on Process cruises has generally been directed at specific questions about the distribution of predators or prey relative to stratification, source and loss regions, or frontal processes. The samples are therefore analyzed so as to address these questions, and not necessarily with the same protocol as the Broadscale samples. In cases where sample analysis is comparable (some MOC-10 hauls), the data will be used to supplement Broadscale data. Sampling on process cruises provides specific information on predator-prey co-occurrence as a function of depth, on size relationships between predators and prey, and is a source of specimens for gut contents analysis and laboratory experiments.

Process MOC-1 samples will continue to be analyzed in Sullivan's lab, and future MOC-10 samples will be sorted in Bollens' lab.

7. Postdoctoral Investigator.

We agreed to advertise immediately for a postdoctoral investigator who will work during Phase III on small predators. The postdoc will be responsible for conducting feeding experiments on process cruises and in the lab.

8. Cruise Plans

Preliminary discussions about cruise plans for 1999 were held with Ted Durbin, Greg Lough and Karen Wishner, followed by more planning on Monday with Dian Gifford and Karen. Our proposal originally included time on 4 process cruises, but budget cuts now limit us to three. During the Durbin cruise in April, we intend to concentrate on the role of small predators, and compare our predation estimates to Calanus mortality measured by Durbin, Runge and Ohman. Our objectives for the cruise in May with Lough are to sample predators of larval fish across the tidal mixing front, using the MOC-10 and LAPIS video systems. The time series analysis of abundance and distribution of hydroids, predators of both copepods and larval fish, will be continued on this cruise, using counts from bongo net tows in the vicinity of the larval fish patch.

On the cruise in June with Wishner and Gifford, we plan to do similar replicated sampling for large and small predators across the tidal front, using the same equipment. It was decided on Monday that three additional days would be needed on the June cruise with Gifford and Wishner to accommodate the MOC-10 sampling. These days may be available from time on Durbin's cruises in March and April.

List of figures

Figure A: Abundance of Clupea harengus on Georges Bank, collected with the MOC-10. Each month represents the mean of all Broad-scale stations sampled. (n=1080 fish)

Figure B: Abundance of Salpa spp. on Georges Bank, collected with the MOC-10. Each month represents mean of all Broad-scale stations sampled. (n=36,093)

Figure C: A representative sample of the amphipod, Themisto gaudichaudi. These predatory amphipods consume copepods when young (1-3mm) and copepods and larval fish when greater than 6mm long. Small juveniles are very abundant (up to 130,000/1000m³) in the upper water column; larger individuals are common in both the upper water column and at mid-depth.