Shallow Water Reef Fish Survey


How is it funded, who administers it?


Why was the survey originally designed?


What are/were the intended outcomes/objectives?

OBJECTIVES Reef fish 88-2002


OBJECTIVES Reef fish 2003-13


Describe the temporal and spatial coverage of the survey

Show the geographic coverage (including depth) of the survey (maps are helpful)


How often is the survey conducted (annually, biennially, biannually etc)?


Is it seasonal?


Describe the underlying experimental design

Fixed site or stratified-random (If stratified-random, describe the strata)?


Describe allocation of sites per strata (proportional/weighted?) and number of sites


Describe the methodology and gear  


Describe the outputs of the survey

Describe the data outputs & include a discussion of the variance structure of the data


Self-Evaluation of pros and cons of survey

What suite of species does this survey target (what are gaps)?


What are the gear/method biases (catchability, size selectivity)?


What are the temporal/spatial limitations?


Is this survey expected to continue into the future?


It is relatively expensive/inexpensive, logistically difficult/easy?


Self-Evaluation of utility of survey for generating information for stock assessment

Spatial/Temporal coverage

Data generated: length frequency, spawning stock biomass, mortality etc.


Provide any other information that may be relevant to this discussion


Provide most relevant documentation (w/PDFs or hyperlinks if possible)


1989 Pilot study used to adopt SEAMAP Methodology


The coastline and shallow water platform area of Puerto Rico west of the 67th parallel was divided into 2 x 2 mile square quadrats and each numbered. (There are approximately 120 such quadrats including areas surrounding  the islands of Mona, Monito and Desecheo).


Of the 120 quadrats, 53 were selected using random number tables. This provided for a coverage of 44% of the entire area.


Over a period of 15 months, each of the 53 quadrats was sampled by trap and hook and line fishing (the predominant methods used by fishermen in these areas) at least two times, giving a minimum of 143 trips in the year. These trips were carried out by two Fisheries Research Laboratory vessels each making a minimum of two trips a week for 40 weeks. Additional trips were made to a randomly selected sub-sample of the 53 originally selected quadrats.


For each trip the following data were recorded:

  1. fishing method (hook and line/fish traps)
  2. date, time spent fishing
  3. weather
  4. quadrat code
  5. physical coordinates and depth
  6. gear effort (i.e., number of traps and duration of    set, number of hooks, number of fishing lines)
  7. total weight of the entire catch by gear type


For each catch (by hook and/or trap) the following data were obtained:

  1. species composition of entire catch
  2. individual size (i.e.,fork length (FL) and/or weight, in grams) for each species
  3. for certain species otoliths were extracted
  4. Sex, and gonadal state of maturation


In order to more closely mimic local fisheries it was decided to used three different sizes of hooks per line (Mustad Sea Kirby Hooks No. 5, 6, and 7). Each line had one of each of the three different hooks approximately one foot apart. Two different kinds of traps were used depending on the depth at which they were set. For depths up to 50 fathom traps dimensions were 4' x 4' x 2', while for depths over 50 fathoms trap dimensions were 6' x 5' x 2'. This decision was taken for practical reasons, since smaller traps are easily carried or moved by strong currents at greater depths. Mesh size of both trap type was 1 1/4". Bait used for hooks was squid and for fish traps tuna gonads were used with which many fishermen bait their traps in the sampled area. To reduce gear loss, traps were set for an average of 6 daylight hours during each trip, while hook fishing was carried out.

Data obtained from the fisheries-independent survey was entered into a micro-computer using DBASE III PLUS and Lotus 1-2-3 to facilitate the preparation of periodic summaries.

Quarterly and annual summaries of monitoring results were made available to NMFS, local fishermen, management agencies, and other interested parties.

Length frequency analyses for those species that comprised the major catch were produced, as well length weight regressions. Outliers and extreme values were removed from length weight regression, visually taking those that were on/or over approximately two standard deviations off the best fitted line.


Reef fish 2003-06



Following the methodology established previously for similar studies (Rosario, 2004), the western platform area of Puerto Rico, up to the 50 fathoms contour, was divided into two squared nautical miles sampling stations (Figure 1).  Mona Island and Desecheo Island were included and their surrounding platform was divided in stations as well.  Each sampling station was defined by four GPS coordinates and identified with a number (Figure 1 and 2; Table 1). Each station was classified according to the following depth ranges:

According to the station bathymetry it could have been classified under the three categories.  Five stations from each category were randomly chosen (Figure 1).  Some stations were added to cover off-shore marine protected areas on the west that has been traditionally monitored (Bajo de Cico, Tourmaline, Abrir la Sierra and Mona Island).  This decision was made after the sampling was begun, reason for which the amount of sampled stations is greater than 15.  



The aim of the study was to sample each station ten times.  The sampled area within a station was only that within the depth range for which the station was chosen (Figures 3-19).  Sampling order was randomly chosen.  Weather conditions, vessel condition and gear and personnel availability determined the sampling dates. 

Two methods were used to sample the shallow-water reef fish at the stations, fish traps and hook and line.  Fish traps were made of 1½” mesh size, with two doors, one of which was tied with a special rope that will deteriorate fast and allow fish to escape in case that the trap gets lost.  Heavier steel rods were used for the trap bottom frame to assure the trap will place itself on the ground with the opening facing up.  A total of fifteen traps were deployed during the sampling.  The traps were divided in five sets of three.  Each trap was identified, at the float, according to their respective set (e.g. 1-1, 1-2, 1-3, 2-1…).  A small cage holding the bait (sardines) was placed at the same side of the trap opening to forbid fish access to the bait without entering the trap.  At each station fish traps were deployed in sets, but individually, at least 150m apart.  Soak time of the fish trap was five hours, from the time when the first fish trap was deployed to the time when the first fish trap was recovered.

Hook and line fishing was performed during fish traps soaking period for four hours.  The vessel was kept adrift, moving it only when the vessel reached the station’s boundaries or areas exceeding the depth range for which the station was chosen.  The vessel was usually turned off while adrift.  The area sampled was determined by recording the coordinates at the beginning and end of the fishing period.  A minimum of three fisher, each using a line with three hooks (size #06 and #04), and using squid as bait. 

Fish collected were placed in bags identified with the date, station number and origin (fish trap ID or fishermen ID).  The fish were taken to the Fisheries Research Laboratory (FRL) facility where they were placed on a freezer until processed.   When processing the fish, it was weighted, measured (total and fork length), sexed and its reproductive stage determined by visual inspection.  Five categories were used for the reproductive stage: undetermined (1), resting (2), enlarging (3), ripe (4) and spent (5).

SEAMAP standard data sheets for every fish trap and fishers were completed.  The data was entered and stored on SEAMAP software 3.0.



As a side study to this project the gonads from a sub-sample of fish caught were collected and preserved for histological analysis.  Pictures of the gonads were taken and identified with the gonad information.  The slides were examined to determine sex and reproductive stage.  The same categories used for the visual identification were used for female fishes.  It was observed that for males, the reproductive stage were not as discrete as in females, for this reason the categories used were mature (M) or immature (I) (Table 11).  The results were compared with the visual classification recorded when processing the fish.  The purpose of this was to create a visual aid for the reproductive stage identification of gonads for the different fish species, and use it as a quality control for the visual identification of the fish gonads.  A total of 84 gonads were collected. Table 11 summarizes the information of the samples collected.  From the collected tissues, 61 were fixed adequately and gave good information.  A selection was made of gonads pictures with their respective histology reading and presented in figure 20.     


2010 Reef fish east and west coasts of PR



The aim of the present survey was to collect, manage, and disseminate fisheries-independent data collection of shallow water reef fish resources and their environment.  These data were used to obtain catch per unit effort estimates, to determine species composition and to evaluate annual trends in the fishery.  The data are also available for comparison with fisheries-dependent data collected under other statistics projects of PR and the USVI. 



Assess the survey design and standardize sampling methodologies identified in the Statistical Survey Design Analysis.  Establish and conduct fishery-independent surveys to obtain CPUE (biomass per unit gear), determine species composition, evaluate trends in the fishery, and characterize the fishery habitats.  Data obtained from the Pilot Study were also analyzed in order to establish the optimal design for the long term Reef Resources Survey.

Project Objective:



  1. Previous 2 x 2 nautical miles quadrants used for the western platform were used.  The east and south coast were divided as well in 2 x 2 nautical miles quadrants. Some details concerning sampling are subject to minor modifications depending on logistics and prevailing conditions of weather and vessels.
  2. Thirty quadrants were randomly chosen for the east and west, and ten for the south.  Quadrants were located by Global Positioning Systems (GPS).  Each quadrant was sampled twice.  Sampling station and date were random selected and varied according to weather and sampling logistics.
  3. At each quadrant fishing was done using hook and line with fish hooks #06, sinker units (weights) and squid as bait. Three lines (equal to three fishers) were used for sampling, each with three hooks. Quadrants were sampled for 4 hours during each trip.
  4. For each trip the following data was recorded:
    1. Date, time (i.e. time out and time returned to dock).
    2. Quadrant code (latitude and longitude).
    3. Fishing time for line to the nearest 15 minutes.
    4. Weather conditions.
    5. Depth.
    6. Total number of hooked fished per vessel.
    7. Number, weight, length, reproductive condition and identification of fish per hook and line as well as by individual fishermen.
    8. Substrate type was characterized whenever possible.
  1. Catches by individual fisherman were kept separated for each fishing trip.  The data was entered with an identification code for each fisher so that it could be analyzed for each fishing member.  These data could provide an estimate of fisherman productivity and also an indication of the variability of individual fisherman performance.
  2. Data was entered, edited, and stored on microcomputer on Access standardized format.  Semi-annual summaries performance and annual reports including data summaries were completed using Excel and Word.  Data was also entered and stored on SEAMAP software and sent to the SEAMAP Database Manager in Pascagoula, MS. Statistical analysis was performed using SigmaPlot 12. The Kolmogorov-Smirnov two-sample test and the t-test were used to compare size frequency distributions and mean size. Sex ratios were tested statistically for significant deviations from the expected 1:1 with a chi-square test (α=0.05) (Sokal and Rolhf, 1981).
  3. A statistical analysis of data, including recommendations on sampling design will follow completion of the study.
  4. Classification of species composition by first, second, third, and trash fish is the general market value presented by Matos and Sadovy (1990) for P.R.  Some modifications have arisen to this classification, as certain species that formerly did not have commercial values are now being reported in landings, with commercial value (Matos, 1993).  This classification varies markedly from coast to coast, but in general, reflects the classification used by the majority of fishermen in P.R.  The two categories that tend to vary most in terms of how species are classified according to their market value are third and “trash” (“brosa”) fish.  The major difference concerns the classification of squirrelfishes.  In certain areas, such as the west coast, this group is considered to have no market value (trash fish); while, in others such as the south coast, it is classified as third class.
  5. Species with a minimum sample size of one hundred individuals for the entire year are taken into consideration for the analysis of length-frequency data.  A 10 mm size class interval is considered the most appropriate for collected groupers, sand tilefish and squirrelfishes.

Histology Procedure:

  1. All fish captured was visually sexed as follows:
    1. Unripe individuals are designated as F1 and M1.
    2. Sub-ripe individuals are classified as F2 and M2.
    3. Ripe individuals are designated F3. (Females with ovaries usually transparent and colourless; enlarge gonad with large, well developed eggs); and M3 (males with testes with loose or running milt).
    4. Spent gonads F4 and M4; individuals with enlarged and flaccid gonads.
  1. Two principal gonad stages were used for each sex to establish the spawning period of selected species: ripe and spent gonads.
  2. All gonads were collected and preserved for histological analysis.  Pictures of the fish and the gonads were taken and identified with the gonad tag information.  The slides were examined to determine sex and reproductive stage. 
  3. The same categories used for the visual identification were used for female fishes.  It was observed that for males, the reproductive stage were not as discrete as in females, for this reason the categories used were mature (M) or immature (I).  The results were compared with the visual classification recorded when processing the fish.  The purpose of this was to create a visual aid for the reproductive stage identification of gonads for the different fish species, and use it as a quality control for the visual identification of the fish gonads. 
  4. Spawning season and size of maturity was calculated for all species with enough data.    



Figure 1. Sampled stations off west coast reef fish survey.


Figure 2. Sampled stations off the east coast of Puerto Rico for reef fish survey 2010.



Figure 3. Sampled stations off the west coast of Puerto Rico during 2013.



Figure 4. Stations sampled during the reef fish survey off the east coast 2013.