320 



Abstract— The objective of this study 

 was to investigate the spatial pat- 

 terns in green sea urchin (Strongylo- 

 centrotus droebachiensis) density off 

 the coast of Maine, using data from a 

 fishery-independent survey program, 

 to estimate the exploitable biomass of 

 this species. The dependence of sea 

 urchin variables on the environment, 

 the lack of stationarity, and the pres- 

 ence of discontinuities in the study 

 area made intrinsic geostatistics 

 inappropriate for the study; there- 

 fore, we used triangulated irregular 

 networks (TINs) to characterize the 

 large-scale patterns in sea urchin 

 density. The resulting density sur- 

 faces were modified to include only 

 areas of the appropriate substrate 

 type and depth zone, and were used 

 to calculate total biomass. Exploitable 

 biomass was estimated by using two 

 different sea urchin density threshold 

 values, which made different assump- 

 tions about the fishing industry. We 

 observed considerable spatial vari- 

 ability on both small and large scales, 

 including large-scale patterns in sea 

 urchin density related to depth and 

 fishing pressure. We conclude that 

 the TIN method provides a reasonable 

 spatial approach for generating bio- 

 mass estimates for a fishery unsuited 

 to geostatistics, but we suggest fur- 

 ther studies into uncertainty estima- 

 tion and the selection of threshold 

 density values. 



Estimating exploitable stock biomass 

 for the Maine green sea urchin 

 (Strongylocentrotus droebachiensis) 

 fishery using a spatial statistics approach 



Robert C. Grabowski 



School of Marine Sciences 



5741 Libby Hall 



University of Maine 



Orono, Maine 04469 



Present address: Flat 6, Falmer House 



16-17 Marylebone High St 

 London, W1U4NY, England 



E-mail address grabowskirc6ig'yahoo com 



Thomas Windholz 



The GIS Training and Research Center 

 Idaho State University 

 Pocatello, Idaho 83209-8130 



Yong Chen 



School of Marine Sciences 

 5741 Libby Hall 

 University of Maine 

 Orono, Maine 04469 



Manuscript submitted 27 January 2003 

 to the Scientific Editor's Office. 



Manuscript approved for publication 



21 December 2004 by the Scientific Editor. 



Fish. Bull. 103:320-330 (2005). 



The green sea urchin (Strongylocen- 

 trotus droebachiensis) is an impor- 

 tant resource of the fishing industry 

 in the State of Maine, where it cur- 

 rently ranks fourth by value. The com- 

 mercial fishing industry began in the 

 late 1980s as a result of expanding 

 foreign markets. Landings reached a 

 peak of more than 22,000 metric tons 

 (t) in 1993. However, declining stock 

 abundances have caused landings to 

 diminish over the last decade, and in 

 2001, less than 5,000 t were landed 

 (Chen and Hunter, 2003). Consider- 

 ing the economic importance of the 

 fishery and its persistent decline in 

 yield, it is essential that we establish 

 an accurate quantitative assessment 

 of the stock in order to develop an 

 effective management plan. 



The Maine Department of Marine 

 Resources (DMR) has collected fish- 

 ery-dependent information since the 

 beginning of the state's commercial 

 fishery. This information, including 

 catch and size-composition data, has 

 formed the basis of most management 

 decisions in the fishery. The fishery is 



currently managed through limited 

 entry, a restricted number of opportu- 

 nity days, and sea urchin size limits, 

 in which legal-size sea urchins have a 

 test diameter between 52 mm and 76 

 mm. The fishing grounds are divided 

 into two management areas based on 

 spatial and temporal variations in 

 spawning (Fig. 1), in which manage- 

 ment differs only by fishing seasons 

 (Vadas et al., 2002). 



Chen and Hunter (2003) conducted 

 the first formal stock assessment for 

 the Maine green sea urchin in 2001. 

 Fishery-dependent data and sea ur- 

 chin life history parameters were 

 used to assess the population dy- 

 namics of the Maine urchin stock. A 

 length-based stock assessment model 

 was used with a Bayesian approach 

 to determine probabilistic estimates 

 of current stock biomass and exploi- 

 tation rate. The study estimated that 

 the current stock biomass was ex- 

 tremely low, about 10% of the virgin 

 biomass. Only fishery-dependent data 

 were available at the time the stock 

 assessment was conducted, but in 



