CRAMP Home

Long Term Monitoring

Status of the Reefs

Site Selection Process

Methods

Diving Protocols
Permits
Site Survey Protocol
Benthic Monitoring
Reef Fish Monitoring
Database
Archiving of CRAMP Survey Data

Study sites

Results

Cooperative Programs

Rapid Assessment

Mapping

Watersheds

Publications



Search the CRAMP Website

Navigate the CRAMP Website

Bibliographic Search

CRAMP Standard Fish Survey Technique

Alan Friedlander

The Oceanic Institute

Makapu’u Point, Waimanalo, Hawai‘i

Introduction

During FY98 initial studies were undertaken to develop a standard fish transecting technique. The initial development effort was conducted on the island of Kaua‘i.

Fish Assemblages

Visual censuses of the fish assemblages associated with benthic habitat transects were conducted using a standard 25m x 5m underwater visual belt transect survey method (Brock, 1954; Brock, 1982). A SCUBA diver swam each transect at a constant speed (~ 15 min/transect), identified to the lowest possible taxon all fishes visible within 2.5m to either side of the centerline (125m2 transect area). Standard length (SL) of fish was estimated to the nearest centimeter. Live wet weight, W, of all fishes recorded in all censuses was estimated from the visually estimated SL using the relation W = a(SL)b. Values of the fitting parameters a and b were derived from previous work of the Hawai‘i Cooperative Fishery Research Unit.

Results

Sample size analysis

An "over sampling" effort of the reef fishes at Ho’ai Bay was conducted on 5 August 1999 to help determine the optimal number of 25m x 5 m transects conducted on that reef. Twelve 25m x 5 m transects were conducted in approximately 10m of water. This study was also conducted to determine the optimal sample size for future work at other locations.

A technique developed by Bros and Cowell (1987) using the standard error of the mean to resolve statistical power was used to determine the number of samples needed based on number of species, number of individuals, biomass, diversity, and evenness. This method uses a Monte Carlo simulation procedure to generate a range of sample sizes versus power. The sample size at which a further increase in sample size does not substantially increase power (decreasing SEM) is taken as the minimum suitable number of samples. A program written by Doug Harper of the NMFS/SEFSC/Miami Laboratory was used to conduct this analysis. For number of species high and low standard error of the mean started to converge at 3 to 5 samples, for number of individuals convergence was achieved at ca. 5-7 transects (Fig. 1). Convergence for biomass occurred at ca. 7-9 transects while the number of samples required for diversity and evenness was between 5 and 7.

Sample size optimization for number of species, number of individuals, biomass, diversity, and evenness. Relationship between standard error of the mean (SEM) and sample size at Ho’ai Bay. Monte Carlo simulation procedure for sample size optimization described by Bros and Cowell (1987).

Figure 1. Sample size optimization for number of species, number of individuals, biomass, diversity, and evenness. Relationship between standard error of the mean (SEM) and sample size at Ho’ai Bay. Monte Carlo simulation procedure for sample size optimization described by Bros and Cowell (1987). (Click for a larger view)

The relationship of sample size with accuracy of the mean was examined for number of species, number of individuals, and biomass from the 12 transects conducted at Ho’ai Bay. Sample means were compared to a theoretical population mean using the t-distribution:

(Eckblad 1991).

The equation was solved for sample size using various values of accuracy (accuracy = [sample mean - population mean]/sample mean). Using a Type I error rate of 0.10, the number of samples needed to detect changes in assemblage characteristic decreases rapidly with relatively slight losses in accuracy. A Type I error rate of 0.10 was chosen over the traditional 0.05 because failing to detect a change when one is actually occurring (Type II error) could lead to population collapse (Gibbs et al. 1998). This is the precautionary approach to management as mandated by the Magnuson-Stevens Fishery Conservation and Management Act.

Estimated number of samples needed to detect changes in the mean A. number of species, B. number of individuals and C. biomass. N = 58, a = 0.10 and 0.20.

Figure 2. Estimated number of samples needed to detect changes in the mean A. number of species, B. number of individuals and C. biomass. N = 58, a = 0.10 and 0.20. (Click for a larger view)

The estimated number of samples needed to detect various levels of change varied greatly among the three parameters (mean abundance of species, individuals, and biomass). Figure 2 provides estimates of the number of samples needed to detect various levels of change in the mean abundance of species, individuals, and biomass. Using a Type I error rate of 0.10, the number of samples needed to detect changes decreases rapidly with only a slight decline in accuracy. Slightly more than one sample is required to detect a 20% change in number of species per census while ca. 8 transects are required to detect a 20% change in number of individuals. Again, biomass is highly variable with ca. 60 samples needed to detect a 20% change. Using a Type I error rate of 0. 20 substantially decreases the number of samples needed to detect change in the 10% to 20% range of accuracy. Slightly more than 1 sample is required to detect a 15% change in number of species per census while 7.7 censuses are required to detect a 20% change in number of individuals. Biomass is highly variable with 35 samples needed to detect a 20% change at Alpha = 0.2.

Detrended correspondence analysis (DCA) of fish transects (25 m x 5 m) conducted at CRAMP survey sites on Kauai in 1999.  Rare species are down weighted.

Figure 3. Detrended correspondence analysis (DCA) of fish transects (25 m x 5 m) conducted at CRAMP survey sites on Kaua‘i in 1999.  Rare species are down weighted. (Click for a larger view)

Comparison of fish assemblage characteristics among CRAMP sites

Detrended correspondence analysis (DCA) was used to identify clusters of similar transects in ordination space. This type of ordination results in an arrangement of samples of species in a low-dimensional space such that similar samples are in close proximity to one another (Gauch, 1982). In this DCA, habitat variables do not influence the ordination; rather, stations with similar assemblage structure cluster together (Greenfield and Johnson, 1990). Transects within sites showed good concordance, and sites appeared to cluster by depth and habitat complexity. The inshore, lower complexity sites tended to cluster in the upper left portion of the figure while the higher complexity, offshore sites clustered towards the lower right portion of the figure. Offshore sites tended to have a wider spread in ordination space due to the greater species richness and diversity associated with these sites.  Fish assemblage characteristics varied among locations and habitats with the Limahuli offshore sites having the greatest number of individuals and highest biomass observed on fish transects around Kaua’i in 1999. 

Comparison of fish assemblage characteristics by location and habitat for CRAMP sites surveyed in 1999. Values are transect (125 m2) means with error bars representing one standard deviation of the mean.

Figure 4. Comparison of fish assemblage characteristics by location and habitat for CRAMP sites surveyed in 1999. Values are transect (125 m2) means with error bars representing one standard deviation of the mean.  (Click for a larger view)

Fish biomass at the Limahuli offshore site was more than twice that observed at the site with the second largest biomass (Ho’ai Bay offshore) and an order of magnitude greater than the inshore habitat at Limahuli. The high spatial heterogeneity of the habitat and the relatively light fishing pressure probably accounts for the high standing stock of fishes observed at this site. High surf during the winter months reduces fishing pressure even further and results in the area being a de facto reserve nearly half of the year. The Ho’ai Bay site in densely populated Po‘ipū had a surprisingly high biomass (16.3 kg/ 100 m2) as well as the greatest species richness observed at any site surveyed.

Fish biomass at the Limahuli offshore site was dominated by large mobile herbivores. Surgeonfishes were the most important family by weight observed at this site, followed by triggerfishes, and parrotfishes. The top five species by weight at this site were all members of the surgeonfish family. These included: Acanthurus leucopareius, A. blochii, A. triostegus, Ctenochaetus strigosus, and A. nigrofuscus, respectively. Nenue (Kyphosus spp.), which tend to feed on drift algae, was the dominant species observed on fish transects at the Ho’ai Bay offshore site accounting for nearly 19% of all fish observed by weight. This species was followed in importance by Chromis ovalis, a planktivorous damselfish that accounted for 15% of the fish biomass at the site. Along the Na Pali coast, the inshore Miloli‘i, Nu‘alolo Kai sites were dominated by surgeonfishes and triggerfishes while on the shallow Limahuli reef flat, small wrasses and surgeonfishes made up most of the fish biomass.

Table 1.  Comparison of fish assemblage characteristics between inshore and offshore habitats at CRAMP study sites around Kaua‘i in June 1999.  Values are grand means per transect.

 

Species

No. of Fish

Biomass (kg)

Diversity

Evenness

Inshore

16.00

137.67

12.57

2.03

0.75

Offshore

25.73

188.96

25.57

2.28

0.70

Overall, offshore habitats had higher values for all fish assemblage characteristics except evenness compared to inshore habitats (Table 1). Offshore biomass was more than twice that observed on transects in inshore habitats while species richness and number of individuals were also much higher offshore. The low spatial complexity and habitat heterogeneity associated with these inshore habitats are the primary reasons for these differences in fish assemblages.

References

AECOS. 1982. Kaua‘i island coastal resource inventory. AECOS, Inc., 970 N. Kalāheo Ave., Suite A300, Kailua, Hawai‘i 96734 report to U.S. Army Engineer Division, Pacific Ocean, Fort Shafter, Hawai‘i 96858. Contract No. DACW84-82-C-0016. 

Bros, W.E. and B.C. Cowell. 1987. A technique for optimizing sample size (replication). Journal of Experimental Marine Biology and Ecology 114, 63-71.

Clark, J. 1992. Beach and ocean recreation study, Ha’ena, Kaua‘i. Division of State Parks, Department of Land and Natural Resources, State of Hawai‘i. Honolulu, Hawai‘i. 49 pp.

Eckblad, J.W. 1991. Biologist’s toolbox: How many samples should be taken. Bioscience 41, 346-349.

Friedlander, A.M. and J.D. Parrish. 1998. Temporal dynamics of the fish assemblage on an exposed shoreline in Hawai‘i. Env. Biol. Fish. 53:1-18.

Friedlander, A.M. and J.D. Parrish. 1998. Habitat characteristics affecting fish assemblages on a Hawaiian coral reef. J. Exp. Mar. Biol. Ecol. 224(1):1-30.

Friedlander, A.M. and J.D. Parrish. 1997. Fisheries harvest and standing stock in a Hawaiian Bay. Fish. Res. 32(1):33-50.

Friedlander, A.M., R.C. DeFelice, J.D. Parrish, and J.L. Frederick. 1997. Habitat resources and recreational fish populations at Hanalei Bay, Kaua‘i. Final report of the Hawai‘i Cooperative Fishery Research Unit to the State of Hawai‘i, Department of Land and Natural Resources, Division of Aquatic Resources. 320 pp.

Friedlander, A.M., J.D. Parrish, and J.D. Peterson. 1995. A survey of the fisheries of Hanalei Bay, Kaua‘i. Final report of the Hawai‘i Cooperative Fishery Research Unit to the State of Hawai‘i, Department of Land and Natural Resources, Division of Aquatic Resources. 87 pp.

Gauch, H.G. Jr, 1982. Multivariate analysis in community ecology. Cambridge University Press, Cambridge. 298 pp.

Greenfield, D.W. and R.K. Johnson, 1990. Heterogeneity in habitat choice in cardinalfish community structure. Copeia, Vol. 4, 1107-1114.

Gibbs, J.P., S. Droege, P. Eagle. 1998 Monitoring populations of plants and animals. BioScience 48 (11): 935-940.

Fish Survey Technique -- Estimating Fish Length

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

808-236-7440 phone

808-236-7443 fax

email: jokiel@hawaii.edu