Zeller et al ; Small-scale fishery catches for US island areas in the Western Pacific 



273 



to-commercial catch ratio of 1.7:1 (Table 3). By the 

 early 1990s, approximately 50% of total catches were 

 thought to be not reported because they constituted 

 noncommercial catches. Thus, the noncommercial catch 

 value for the time period 1993-2002 was set equal to 

 the total commercial catches (Table 3). Thus, we as- 

 sumed higher reliance on noncommercial fishing in the 

 early 1980s compared to the 1990s. We interpolated 

 the proportion of noncommercial catches between 1984 

 and 1993 and expanded them by using reported com- 

 mercial catches. 



Catch rates Re-estimated catches were converted to per 

 capita catch rates by using human population census 

 data, and to catch per unit area of the depth-defined 

 potential coral reef ecosystem habitat area isensu Rohm- 

 ann et al., 2005). Total potential coral reef area to a 

 depth of 100 fathoms (183 m) for CNMI is 476 km^ 

 (Rohmann et al, 2005). Given that most fishing in CNMI 

 occurs near the three main islands, the coral reef area 

 estimate for these islands (331.2 km-) was used here 

 also (Rohmann et al., 2005). 



American Samoa 



Total catches for nopelagic species for American Samoan 

 have been re-estimated independently by Zeller et al. 

 (2006a), and are summarized by decade in Table 4. 

 American Samoan catches were included in the pres- 

 ent study for completeness in the re-estimation of total 

 time series catches for the U.S. flag-associated Pacific 

 island areas. 



Results 



The catch re-estimation for nonpelagic species for the 

 major U.S. flag-associated island areas in the western 

 Pacific combined (excluding Hawaii) indicated two main 

 points (Fig. 2A); 



1 a substantial discrepancy between officially re- 

 ported catch data and potential total catches as re- 



estimated here and by Zeller et al. (2006a). For 

 the time period for which reported data existed 

 (1965-2002), such data may have yielded an under- 

 estimate of likely total catches by as much as a factor 

 of 4.55. This discrepancy was largest in early years; 

 and 

 2 a potential decline of 77% occurred in total catches, 

 from an estimated 2165 t in 1950 to 496 t in 2002. 

 This decline contrasted with the trend observed 

 from the data reported by individual island enti- 

 ties — namely an increasing trend from 147 t in 1965 

 to 269 t in 2002. 



Individual Islands 



For Guam, the re-estimation indicated a decline of 86% 

 in catches of nonpelagic species over the 50-year time 

 period considered here. There was also a 2.5-fold differ- 

 ence between the re-estimated catches and the reported 

 statistics for the 1965-2002 period, driven by under- 

 reporting of catches in earlier periods. Guam's ongo- 

 ing commitment to and consistent application of creel 

 surveys to estimate total catches has resulted in what 

 may be the most reliable estimates of total catches for 

 any of the islands considered here, at least since the 

 mid-1980s (Fig. 2B). Based on the re-estimated data, 

 the annual per capita catch rates for Guam's coral reef- 

 and bottom-fisheries may have declined from 16.0 kg to 

 0.8 kg between 1950 and 2002 (Table 5). Catch rates per 

 area of potential coral reef habitat (to 100 fathom=183 m 

 depth) appear to have declined from 4.7 t/km^/year to 

 0.6 t/km2/year between 1950 and 2002 (Table 5). 



For CNMI, the re-estimated catches indicated a de- 

 cline of about 54% in catches of nonpelagic species be- 

 tween 1950 and 2002. Comparing the catches reported 

 by CNMI from WPacFIN with the re-estimated total 

 catches, we found a 2.2-fold under-reporting of poten- 

 tial total catches by the reported data, compared to the 

 re-estimated totals for the 1983-2002 time period of 

 coverage by WPacFIN (Fig. 2C). Taking into account 

 CNMI's rapid human population growth over the last 

 two decades, we surmise that the annual per capita 

 catch rate may have declined from a high of 72.6 kg in 



