In:
Canadian Journal of Fisheries and Aquatic Sciences, Canadian Science Publishing, Vol. 68, No. 5 ( 2011-05), p. 880-897
Abstract:
Conservation objectives aimed at maintaining the diversity of Pacific salmon ( Oncorhynchus spp.) are often expressed as a desire to ensure that spawner abundance is spread out over a number of spawning sites. However, sampling is not usually possible at all sites or in all years. For such incomplete data sets, rotating panel sampling designs and hierarchical estimation models have been suggested as ways to improve monitoring performance. To evaluate the potential benefits of using these approaches to assess the conservation status of coho salmon ( O. kisutch ), we developed a simulation procedure that modelled spatial and temporal variation in salmon abundance at multiple sites within a region. Results show that both approaches were largely unsuccessful at reducing classification errors for conservation status. Furthermore, indicators describing distributions of abundance levels and temporal trends in abundance within a region were more sensitive to missing data than to observation error variance on annual abundance estimates. Thus, sampling effort might be better spent reducing the level of missing data within a regional data set, as opposed to obtaining more precise estimates for only a few site–year combinations. Our results also show that the best monitoring plans for regions depend on monitoring objectives as well as the relative magnitudes of spatial and temporal variability.
Type of Medium:
Online Resource
ISSN:
0706-652X
,
1205-7533
Language:
English
Publisher:
Canadian Science Publishing
Publication Date:
2011
detail.hit.zdb_id:
7966-2
detail.hit.zdb_id:
1473089-3
SSG:
21,3
SSG:
12
