Given Location of Midwater Trawl Catch, Hindcast Simulation Suggests Likely Origins of Larval Shorefishes

To effectively achieve conservation goals and regulate human activities in marine ecosystems, resource managers need to understand the connectivity of fish and other biota in the ecosystem—how organisms in a particular area or habitat are related to those in a different location through colonization and movement in a spatially-structured environment. When marine scientists consider the spatial distribution of reef fishes, they find that most adult reef fish show site fidelity (limited movement), thus dispersal within the fish population is limited to the mobile larval stage of the fish. Gathering information on dispersal of these organisms using a vertically stratified sampling scheme is intractable as the larvae are not easily captured, occur in low densities, and are spread over a large geographical area.

To better understand the distribution of larval shorefishes, Fishery Biologist Don Kobayashi of the PIFSC Ecosystem and Oceanography Division is teaming with University of Hawaii oceanography graduate student Johanna Wren. Together, they have been analyzing catch data from Cobb midwater trawl sampling conducted by PIFSC scientists and collaborators during research cruises of the NOAA Ship Oscar Elton Sette in 2011 (cruise SE-11-04) and 2012 (cruise SE-12-06). With details on the fish catch location in hand, the scientists are using a hindcasting dispersal model to predict the likely origination site and depth stratum occupied by the larval reef fish. The dispersal model is driven by HYCOM (HYbrid Coordinate Ocean Model, data from University of Hawaii) which projects ocean currents backwards in time. Model projections originate at the trawl site and extend back in time up to 75 days prior to the trawl date. The investigators hope to predict the likely origination site and depth strata of occupancy for larval shore fish. This information will improve understanding of horizontal and vertical movements of shore fish larvae and will be useful for developing ways to sample these elusive life history stages.

In their research, Kobayashi and Wren are trying to determine what depth, or range of current depths, the shorefish larvae occupy while dispersing. By comparing five different depth scenarios (surface waters, 50m depth, 100m depth, 0-50m average, 0-100m average) and by using a plausible range of pelagic larval durations (15, 30, 45, 60, or 75 days), they hope to predict origination sites and determine which depth layer is the optimal and most likely dispersing depth of the larvae.

Figure 1. Hindcasted sites of origin for larval reef fish caught at various trawl locations under 5 assumed vertical strata for larval dispersal within currents.  Results illustrate differential geographic transport and success of dispersing larvae. From Oscar Elton Sette cruise SE-11-04.
Figure 1. Hindcasted sites of origin for larval reef fish caught at various trawl locations under 5 assumed vertical strata for larval dispersal within currents. Results illustrate differential geographic transport and success of dispersing larvae. From Oscar Elton Sette cruise SE-11-04.

For larvae captured in trawl hauls off the Big Island of Hawaii, we see there is considerable variation in the geographical distribution and areal expanse of the hindcasted sites of origin depending on the assumed depth stratum used by dispersing larvae (Figure 1.) Assuming dispersal in the 50 m depth stratum, hindcast simulations indicate the likely origin sites are located mainly on the northeast and southeast coasts of Hawai‘i Island. At the 100m depth, the pattern changes and the main sites of origin are predicted to be located on the Kona and southeast coast of the island of Hawai‘i and throughout the Maui Nui complex (Maui, Moloka‘i, Lāna‘i and Kaho‘olawe). The two average depth scenarios indicate origin sites on the windward coast of Hawai‘i. In the deeper strata, hindcasted sites of origin are in the Maui Nui complex. Thus, in the averaged depth scenarios for larval dispersal, it appears that the shallower part of the depth range has a greater influence on determining sites of origin.

Figure 2 shows simulated trajectories of hindcasts for larvae from a subset of the trawls made off the leeward coast of the Big Island assuming larval dispersal in the 100 m depth stratum. In the maps, simulated location points are color-coded to indicate backward movement in time. In this example, the only predicted site of origin for larvae caught in these particular surface trawl catches is located on the windward coast of the island of Hawai‘i.

 Figure 2. Hindcasted dispersal trajectories of larval reef fish caught at SE-11-04 trawl sites off the leeward coast of the Big Island of Hawaii assuming dispersal in the 100-m depth stratum. Color indicates number of days into the past from the time of catch in the trawl (dark blue) to 75 days earlier (dark red).
Figure 2. Hindcasted dispersal trajectories of larval reef fish caught at SE-11-04 trawl sites off the leeward coast of the Big Island of Hawaii assuming dispersal in the 100-m depth stratum. Color indicates number of days into the past from the time of catch in the trawl (dark blue) to 75 days earlier (dark red).