The collection and accurate interpretation of both fisheries dependent and fisheries independent data are of fundamental importance to our understanding of fished species. Both should be viewed together in order to gain a full understanding of the nature of changes in the magnitude of landings and productivity of the resource.

Fisheries dependent data include landings, port sampling, and sea sampling data gathered in conjunction with harvesting. Such data are influenced by many variables specific to how fishermen harvest their catch, including: area fished, number of fishermen, intensity of fishing effort, gear specifications, level of expertise of the fishermen, and the availability of the fish (or lobster).

Fishery dependent information is important in its own right. One of the keys to understanding changes in the various indices of abundance is an accurate accounting of the removals. Part 2: Fisheries data collection Some removals are landed, while others occur incidentally to fishing activity.

Fisheries independent data include surveys that are not directly influenced by harvesting activities. Let’s look at a demographic example to illustrate the significance of sampling design.

Say you were interested in estimating the density of people in the US. You would want to make sure that sampling was done in all regions and perhaps stratified into cities to allow for valid extrapolation to a total.

Two very different pictures would emerge if all of the samples happened to fall in New York City or if all the samples happened to fall somewhere in North Dakota.

To protect against this unfavorable outcome, it is necessary to implement a design in which each segment of the resource has an equal chance of being sampled.

The power of any survey-based independent estimate is that you can make a valid inference about the unsampled fraction of the resource. Over the long haul, the proportion of stations with and without fish (or lobsters) in the sample should approximate what is found in the population.

The sample taken in New York City cannot be extrapolated to the total because the high-density area was chosen in advance and based on some prior knowledge. Taking additional samples in New York City would not improve the estimate of the total. At the same time, to extend and mix the demographic metaphor, fishermen cannot afford to fish in North Dakota, but must fish in New York City.

In the context of scientific sampling design, fishery dependent information is important because fishermen build up a working knowledge of the resource that can help refine sampling strategies and improve precision.

In human populations, information about city populations is a crucial component in the design of sampling strata and in the allocation of samples to strata. In the same way, fishermen’s knowledge is critical to refining survey strata and sample allocation for fisheries independent data.

Acknowledgment: Special thanks to Dr. Paul Rago of the Northeast Fisheries Science Center for critical comments on a draft of this essay and valuable input on this version.