Application of an experimental approach to management of a tropical multispecies fishery with highly uncertain dynamics

In m ost m ultispecies fisheries th ere is considerable uncertain ty in selection o f an app ro p ria te m odel to rep resen t the dynam ics o f the resource. M any m odel structures and /o r p aram eteriza tions m ay be consistent w ith ecological princip les and the avail­ able da ta . T hese m odels m ay have very d ifferen t m anagem ent im plications, bu t may be im possible to distinguish by p rocess-orien ted research at reasonab le cost and on a tim e fram e of relevance to m anagem ent. In som e circum stances an adaptive experi­ m ental m anagem ent regim e can be econom ically beneficial by allow ing em pirical learn ing abo u t resource dynam ics and d iscrim ination betw een alternative m odels. H ow ever, in any particu lar situation it m ust be d eterm ined w hether an experim ental regim e is econom ically viable, and which m anagem ent actions and research obse r­ vations should be included in the regim e. T he developm ent and application o f an experim ental m anagem ent regim e for the fisheries opera ting on a trop ical fish com m unity in northw estern A ustra lia are described . T h e h istory o f exploita tion is sum m arized and a num ber o f sim ple m odels are suggested which can m im ic past changes observed in fish com m unity com position. T hese m odels include interspecific, intraspecific, and hab ita t m odification m echanism s. Possible socio-econom ic re ­ sponses of the fishing industry to changes in th e resource sta te are im p o rtan t to evaluation o f a p rospective fishing regim e, and these are also m odelled . T h e m odels are used to evaluate options fo r m anagem ent o f com peting trap and traw l fisheries on the N orthw est Shelf. It was found th a t if an experim ental m anagem ent regim e w ere adop ted fo r abou t 5-15 years (during which tim e key uncertain ties in the resource dynam ics and socio-econom ic responses could be resolved) a larger expected value could be ob ta in ed from the resource than if the existing m anagem ent regim e w ere continued. Som e experim ental m anagem ent regim es also prov ided a g rea te r expected value than would be ob ta ined from im m ediate application of the m anagem ent regim e th a t is optim al fo r any of the individual resource m odels. E xperim ental m anagem ent periods o f less than abo u t 5 y d id n o t allow sufficient reso lu tion o f uncerta in ties to be w orthw hile, and periods o f longer than about 15 y o ften resu lted in the costs o f ob tain ing th e additional resolu tion exceeding the value o f the expected im provem ent in re tu rn s from the resource.