MRFSS Surveys

by Thomas Siciliano

Every year for the past six years the National Marine Fisheries Service has held a Constituent Data Review Workshop. I have attended the last two years and just returned. The goal of the meeting is to preview the 2008 catch and effort estimates, understand how they were derived, and discuss any specific concerns or questions regarding the 2008 data collection effort and the resultant estimates.

I was hopeful that there would be sufficient detail to understand how the numbers are derived. Dave Van Voorhees gave an overview. Gordon Colvin talked about the new Marine Recreational Information Program and Dr. Ronald Salz discussed converting survey data to catch estimates.

The MRFSS survey consists of two components, the intercept survey which is the catch data used and the random household telephone survey which defines the effort. The data from the two methods are combined by some mysterious method to estimate the number of anglers and fishing days. Even though the document is over an inch thick, the best that it can be described as is an executive overview. There is very little detail to truly understand how the numbers, which are used by fisheries managers to set the quotas that we have struggled with, are derived.

The success rate of the phone survey in finding a family that has an angler who has fished in the last two months is in the range of 1%. The new registry should improve this ratio immensely. My guess is that it should improve to at least 50% and for some waves could reach 90%. This will greatly help the effort side of the data collection. Nowhere in the report is the number of phone calls mentioned.

Page 15 of the report states that the quality of the estimates is largely dependent on sample size. I asked if the new survey would also increase the dockside intercepts by a percentage similar to that expected from the phone survey. I was shocked when the answer came back as maybe they would double the dockside surveys. I asked the question differently of another speaker later in the program. I asked him to make the assumption that funding was not a constraint and they could get all the money they needed to do the job right. I could not get an answer.

The MRFSS conducted 1850 angler interviews of the private/rental boat for the coast for the year for all species. The JCAA fluke tournament survey data provides more hard numbers in terms of catch and effort in one day than the entire intercept survey does for the year. In 2008 JCAA had 653 boats with an average of three anglers reporting their catch and effort information. In one day JCAA generated approximately the same amount of data that was used by MRFSS for the entire year of 2008 to determine the catch and effort for all fishing in New Jersey.

An apparent flaw in the data presented is that unsuccessful trips are not presented. In my mind I would think that this would be an important piece of data and should be included in any report.

Here are some observations of the data in the report. The number of angler trips for the last 10 years are presented on a graph. For the Mid-Atlantic region the data shows a low of 14 million trips in 1999 up to a high of 23 million trips in 2007. The number of trips jumps up and down and no explanation is given to the variation. There is no attempt to verify the number with bait sales, party boat trips or gasoline prices. In 2008 the data shows about 12% less trips than 2007. I guess people who lose their house, lose their jobs and have their cars repossessed must go fishing. The New Jersey data shows 9% less trips taken in 2008 than in 2007.

Frustration with the survey data is not confined to New Jersey. The people from Maryland did an analysis of the shore catch of summer flounder, which clearly showed what can happen when scant data is multiplied by the total number of anglers in the state. Maryland has a minimal summer flounder fishery. There is only one place in Maryland where you might have a chance to catch one from shore. It is a bridge on Route 50 in Ocean City. Based on two intercepts (2 fish) it was estimated that 14692 keepers were caught in September 2008 from shore. The Maryland number is 54 fish for September from shore. What is missing is the unsuccessful trips and expanding the data to areas where no fish are caught.

Florida has shown a steady increase in angler trips over the years. In 2008 the West Coast had an increase, yet the East Coast showed a drop of 25%. Inconsistencies like these don’t instill confidence in the minds of recreational anglers. NOAA has a long way before anglers will believe the statistics!

Question for readers. Why does the United States do a census every 10 years? Why don’t they just do a statistical survey of say 1% of the population? The statistics should be pretty good at a 1% sampling level, shouldn’t they? They don’t sample because they want the best possible data, so they do a census. Now, I’m not suggesting that NFMS do a census of all recreational anglers. I would suggest that given the complexity of fishing data the sampling rate should be much higher than it currently is. Let’s look at the party and charter boat information. NMFS conducts a for-hire survey even though they currently have census data on all federally permitted party and charter boats. In 2008 NMFS conducted 2448 interviews. Given that the party and charter boats account for approximately 8% of the angler trips, that is a level of 0.58 %. I wonder how the survey data compares to the census information that has been submitted for the last 20 years? The survey rate for private and rental boats is 0.05% and for shore anglers even less at 0.03%

I would like to put out a call to any statisticians to join a committee to come up with an appropriate number of intercepts they feel would be the minimum number necessary to give or obtain accurate catch numbers.

[News Contents] [Top]