The AGRRA protocol for fishes includes two distinct methods that should both be applied at each site. All transects used for fish assessment should be located within the same depth intervals specified for the benthic assessment (1-5 m and 8-15 m). Transects for fish will tend to be further apart and may range deeper and shallower than transects for benthic organisms. The integration of fish and benthic sampling, while beneficial, will require close coordination between team members for the two parts. It is recommended that the fish observations be conducted between 1000 and 1400 hours if at all possible, when visibility underwater is at a maximum due to overhead sunlight. Many fishes are wary of humans, hence it is necessary to keep away from other people while making these observations.

Method I:       Belt transect counts for defined species list.

1.                   For each transect, record the following information: recorder's name, date, time of start of transect, site name, site number, latitude and longitude, transect #.

2.         Lay a 30 m transect line by first placing the weighted end of the line on the bottom, and then swimming in a straight line while releasing it from the reel as you count the fish. This minimizes the disturbance to the fishes prior to their being counted. Periodically fixing on an object in the distance as you swim will help you swim in a straight line. (You can clip the transect tape to your weight belt to allow for easy release of the tape).

3.         As you swim out the full 30 m transect line, count and record fish found within a 2 m wide visually estimated belt transect. Carry a data sheet in standard format, and a 1-m wide T-bar to ensure accurate monitoring of the 2 m wide belt. Hold the T-bar ahead of you angled downward at about 45 degrees, and try to focus your gaze on the several meters of the transect ahead of the T-bar. Count only those SPECIES listed below and do not count juvenile parrotfishes or grunts less than 5 cm in total length. This list of species has been chosen to provide coverage of a number of the species most likely to be affected by human impacts, while preserving a relatively consistent search image. This should enhance the precision of transect data.

SELECTED FAMILIES- include EVERY SPECIES within the following families:

Acanthuridae (Surgeonfish) (e.g., Acanthurus bahianus, A. chirurgus, A. coeruleus)

*Balistidae (Triggerfish) ONLY Balistes vetula (queen triggerfish), B. capriscus (gray triggerfish), Melichthys niger (black durgon), Aluterus scriptus (scrawled filefish), Cantherhines pullus (orangespotted filefish), C. macrocerus (whitespotted filefish)

Chaetodontidae (Butterflyfish) (e.g., Chaetodon capistratus)

Haemulidae (Grunt) (e.g., Haemulon flavolineatum, H. chrysargyreum, H. sciurus, H. plumieri)

Lutjanidae (Snapper) (e.g., Lutjanus griseus, L. apodus, L. mahogoni, Ocyurus chrysurus)

Pomacanthidae (Angelfish) (e.g., Pomacanthus paru, P. arcuatus, Holocanthus tricolor)

Scaridae (Parrotfish) (e.g., Sparisoma viride, S. aurofrenatum, Scarus taeniopterus, S. vetula)

*Serranidae (Groupers), But Only Genera: Epinephelus and Mycteroperca spp. (e.g., Epinephelus guttatus, E. fulvus, E. striatus, Mycteroperca bonaci)

DO NOT COUNT Haemulidae (grunt) or Scaridae (parrotfish) fish LESS THAN 5 cm in length.

ALSO COUNT the following five species:

Bodianus rufus            (Spanish hogfish)

Caranx ruber (Bar jack)

Lachnolaimus maximus (Hogfish)

Microspathodon chrysurus (Yellowtail damselfish)

Sphyraena barracuda (Barracuda)

*For these families, only count those genera or species listed.

4.         Estimate the size of each fish and assign them to the following size categories (<5 cm, 5-10, 10-20, 20-30, 30-40, >40cm) using a 1 m T-bar with 5 cm increments to assist in estimating sizes. Large groups of individuals of a species will be classified by attempting to put them into one or more size categories as necessary. By remembering to keep effort equivalent on all segments of the transect, you can limit the tendency to count all members of a school crossing the transect, instead of just those members which happen to be within the transect as counting of that segment takes place.

Note:  Sample the transect belt giving uniform attention to each successive 2-m segment. This requires swimming at a more or less constant rate, while looking consistently about 2 m ahead of your current position. You may pause while recording data, and then start swimming again. It is important to swim in a uniform manner. A speed that covers each 30-m transect in 6-8 minutes should be attempted. High densities of counted species will slow this rate in some cases. You will be tempted to count all members of a school as they swim across your transect unless you concentrate on giving equal effort to sampling each successive portion of the transect. Only those school members that are actually within the 2 m wide strip of that segment of the transect at a given time is included in the census. Fish observers should be trained to estimate fish lengths by using consistency training methods both on land and underwater.

5.         When you reach the end of the transect line, stop the survey and recoil the transect tape.

6.         Continue conducting haphazardly-positioned 30 m transects at least 5 m laterally away from the previous position. Repeat the above steps for each transect.

7.         Conduct a minimum of ten (10) transects at each site and one calibration standard (see Standardization section).

Modifications:  Some workers may want to census other species of fish. This is encouraged, provided that these other species are counted on a SEPARATE pass over the transect, after the AGRRA run. Otherwise the census method is substantially changed, and your data may not be directly cross-comparable with other AGRRA assessments.

Method II.      Rover Diver census

After finishing the belt transects (or concurrently depending on the number of surveyors), conduct a roving diver census of ALL SPECIES of fishes following the methodology of Reef Environmental Education Foundation (REEF) (http://www.reef.org/) and briefly explained below. See Scmitt et al (1998) for more details.

1.         The Rover diver census is conducted in the same general area as the belt transects are set.

2.         Swim around the reef SITE for approximately 30 minutes (preferably 45-60 min) and record ALL fish species observed. Use all knowledge you have of fish habits, and search under overhangs, in caves, and so on.  The objective is to find the maximum number of species that you can in during your search time.

3.                   Estimate the density of each species by using logarithmic categories: Single (1fish), Few (2-10 fishes), Many (11-100 fishes), or Abundant (>100 fishes).

4.                   Record your observations on the standardized REEF data entry sheet.

5.                   Submit data to REEF database.

OPTIONAL COMPONENTS

Several other useful assessments may be easily integrated into the core portion of the protocol given above. These optional components, while not part of the core methods, can yield additional information that may lead to a better understanding of the condition of a reef. These optional components include coral recruitment (additional sampling effort), fish bites, and the stationary method for assessing fish communities.      

Coral Recruitment:  Coral recruitment is an important indicator of a reef's regeneration potential and is NOW incorporated into the standard AGGRA protocol. If you would like to increase your sample size of coral recruits, we suggest the following:

            Methods:

1.                   After you have completed your 6 standard benthic transects (if time permits), swim in a haphazard fashion around the reef being surveyed and place the 25 x 25 cm quadrat on the substratum in areas lacking large (> 25 cm diameter) sessile invertebrates.

2.                   Count all small (maximum diameter 2 cm), stony corals (scleractinians and Millepora) that you can see within the 25 X 25 cm quadrats.

3.                   Whenever possible, record their scientific names at least to the level of genus.

4.                   Try to repeat for a total of at least 80 squares (an overall sample of 5 square meters of reef surface).

Note:  Proper training and good eyesight (or corrective lenses) are essential to accurately detect the presence of small corals due to their inconspicuous size and nature.

Herbivory:  The objective of the Fish Bite Method (Steneck 1985) is to gauge the effect of herbivorous fishes on algal composition by quantifying their level of herbivory. Fish herbivory is assessed by counting the observed number of bites per square meter of different guilds of herbivorous fishes, which are categorized as: 

Scrapers =  Scaridae (parrotfish)

Browsers = Acanthuridae (surgeonfish), Microspathodon chrysurus (yellowtail damselfish)

Non-denuders = other Pomacentridae (damselfish)

       Methods:

1.         Use the 1 m line in conjunction with natural landmarks on the reef surface (e.g., a small coral or gorgonian) to haphazardly delineate an area that is approximately 1 m square and representative of the benthic cover on the reef substratum. (Please do not place a meter quadrat to mark your observation area, as some fish are particularly prone to biting novel objects placed within their feeding territories).

2.         Back off as far as you can while still seeing the meter square area. Watch for 5 minutes. Record the depth, time of day, and number of bites from all species of fishes in the three guilds listed above (identify them to species as best as you can). Repeat for a total of 5 squares (and ~25 minutes of observation).

Note:  You must be able to distinguish (a) juvenile scarids from other fishes with similar stripes, such as acanthurids and labrids (wrasses- which only look as though they are biting algae as they search for amphipods to eat) and (b) yellowtail damselfish (which are browsers) from other species of damselfish (that cultivate algal gardens). Be sure to remember to record the time of day since fish activity varies temporally.

The Stationary Plot Fish Survey: The Stationary Plot Technique (Bohnsack and Bannerot 1986) has been used extensively in the Caribbean, particularly the Florida Keys. It has provided data on abundance and size for a wide range of species. We encourage workers to use this method as a third way to quantify fishes at each site, but not as a replacement for Method I or II.

Methods:

1.         Count the number of fish you observe in a visually estimated 7.5 m radius area for a period of 5 minutes.

2.         Estimate and record the lengths for each fish counted.