CRAMP Rapid Assessment. Fish Survey Techniques
Fish populations are highly variable, requiring numerous transects to quantify absolute values of fish communities. Spatial and temporal variability can reduce statistical power by increasing standard deviations. The rapid assessment technique (RAT) was designed to use quantitative, relative values to compare stations and sites relative to others. This can be calculated within a site, by island or statewide. In this manner, RATs can cover a large spatial region and keep costs and effort at a minimum, while maintaining statistical integrity by developing a large sample size. In addition, its design allows for statistical comparability with the more intensive, repeatable Long-term monitoring transects.
To encompass as wide a spatial range as possible and to address the issue of spatial variability, a many but small sampling strategy was adopted (McCune and Lesica, 1992). The RAT is a trade-off between size and number of sampling units. This technique provides an efficient sampling design to assess extremely large areas. There are many advantages to selecting many, short transects over fewer transects of longer length (McCune and Grace, 2001).
Fish populations are quantified using standard visual belt transects (Brock 1954). Transect location is determined using pre-selected random points. SCUBA divers swim along one 25 m x 5 m transect (125 m2) at each station recording species, quantity and total fish length. All fishes are identified to the lowest taxon possible.
Kuulei Rodgers recording fish numbers and lengths on rapid assessment at Waikīkī in February 2002. Photo by Erica Muse.
Total length is estimated to the nearest cm in the field and converted to biomass estimates (tons/hectare) using length-weight fitting parameters. In order to estimate fish biomass from underwater length observations, most fitting parameters were obtained from the Hawai‘i Cooperative Fishery Research Unit (HCFRU). Additionally, locally unavailable fitting parameters were obtained from Fishbase whose length-weight relationship is derived from over 1,000 references. Congeners of similar shape within certain genera were used in those rare cases lacking information.
Conversions between recorded total length (TL) and other length types (e.g. fork length FL) contained in databases involved the use of linear regressions and ratios from Fishbase linking length types. A predictive linear regression of logM vs. logL was used in most cases to estimate the fitting parameters of the length-weight relationship. Visual length estimates were converted to weight using the formula M = a * Lb where M=mass in grams, L=standard length in mm and a and b are fitting parameters.
Any anomalous values were detected by calculating a rough estimate for a given body type. The general trend for a 10 cm fish of the common fusiform shape should be approximately 10 g. Gross deviations were replaced with values from the alternate source.
Trophic levels for fish species were determined using published Fishbase data. The trophic categories included: piscivores, herbivores, detritivores, mobile and sessile invertebrate feeders, and zooplanktivores.
To minimize observer variability, only two divers were used in fish assessments. Calibration of the divers was conducted at Kahe Point, O‘ahu (four transects) and Puhi Bay, Hawai‘i (eight transects). No significant differences were found between the two divers for estimates of number or length of fishes.
Last Update: 04/21/2008
By: Lea Hollingsworth
Hawai‘i Coral Reef Assessment & Monitoring Program
Hawai‘i Institute of Marine Biology
P.O. Box 1346
Kāne‘ohe, HI 96744