SIMULATION SETUP Modeled after conditions found in the central coast of California (CalCOFI line 70) during a typical July Domain is unstructured in alongshore direction Forced by stochastic wind and alongshore pressure gradient, derived from observation Larvae are assumed to follow surface water parcels and are released daily for 90 days within the inner 20 km. Settlement occurs after the larvae have developed competency for their next life stage and if they find suitable habitat We consider settlement successful if a larva is found within the inner 20 km within 20 to 40 days after release Biological sources of larval mortality are not included CONNECTIVITY AMONG NEARSHORE ECOSYSTEMS: NATURE OF LARVAL TRANSPORT S. Mitarai 1, D.A. Siegel 1, C.J. Costello 1, S.D. Gaines 1, B.E. Kendall 1, R.R. Warner 1 and K.B. Winters 2 1 Institute for Computational Earth System Science, University of California, Santa Barbara, Santa Barbara, CA Integrative Oceanography Division, Scripps Institution of Oceanography, La Jolla, CA ABSTRACT Key to the predictive understanding of many nearshore marine ecosystems is the transport of larvae by ocean circulation processes. Many species release thousands to billions of larvae to develop in pelagic waters, but only a few lucky ones successfully settle to suitable habitat and recruit to adult life stages. Methodologies for predicting the larval transport are still primitive, and simple diffusive analyses are still used for many important applications. In this study, we investigate mechanisms of larval transport using idealized simulations of time-evolving coastal circulations in the California Current system with Lagrangian particles as models for planktonic larvae. Connectivity matrices, which describe the source- destination relationships for larval transport for a given larval development time course, are used to diagnose the time-space dynamics of larval settlement. Many important fishery management applications require knowledge of fish stocks on a year-to-year or generation-to-generation basis. For these short time scales (typically less than 1 year), larval dispersal is generally far from a simple diffusive process and the consideration of the stochastic and episodic nature of larval dispersal is required. This work provides new insights into spatial temporal dynamics of nearshore fish stocks. GOALS OF THIS STUDY Understand the role of larval transport in predicting nearshore fish stocks & its proper management Investigate source-to-destination relationships for Lagrangian particles as models for planktonic larvae which originate & settle in nearshore environment by using idealized realizations of coastal circulation tied to real data Develop simple models for the connectivity of nearshore habitat based on the obtained simulation results, and use them in fish stock/harvest models Figure 1: Depictions of the sea level distribution (color contours in cm) and sample larval trajectories. The circles show the location of the larvae while the white trails behind each show their previous 2 day trajectories. The vertical dashed red line indicates the boundary for the nearshore habitat where from which larvae are released from and settlement can occur. SUMMARY & FUTURE PLAN Summary The coastal circulation simulation suggest that the connectivity of coastal habitats is heterogeneous when viewed on short time scales (e.g., less than 10 years). Such stochastic will create unavoidable uncertainty in fish recruitment, potentially complicating the management of nearshore ecosystem. We proposed a new simple model that reasonably accounts for stochastic nature of larval transport found in the simulations. Future plan Test more elaborate larval behaviors and address its role in the connectivity of coastal habitats. Assess the role of headlands Use the proposed model in fish stock/harvest model in place of a simple diffusion model. STOCHASTIC NATURE OF LARVAL SETTLEMENT Only a limited number of independent settlement events occur during one spawning season Sometimes, arrival events occur coincident with reversals in the alongshore wind, which would advect surface water parcels onshore. More often than not, successful settlement occurs because eddy advect larvae to suitable habitat Figure 2: Time series showing a) the departure density of successful settlers at a given alongshore location (vertical axis) and at a give time (horizontal axis), b) the arrival density of successful settlers at a given alongshore location and at a given time and c) the alongshore wind speed forcing the model. The first larvae are able to settle on day 20 (left vertical dashed line). Larval releases stop at day 90 (middle vertical dashed line) and settlement is possible until day 130 (right vertical dashed line). Consider connectivity among nearshore sites as a superposition of “larval packets” Describe the number of “larval packets” based on simple scaling N = (T/t)(L/l)f, where T is larval release duration, t is Lagrangian de-correlation time scale, L is the domain size, l is Rossby radius of deformation (or the packet size) and f is larval survivability. Source and destination of each larval packet are randomly assigned based on the distribution given by diffusion model Nearshore eddies sweep larvae together into “packets” which stay together through much of pelagic stage CONNECTIVITY AMONG NEARSHORE HABITATS Connectivity is heterogeneous & intermittent even in unstructured domain Connectivity can differ for differing life histories Connectivity can differ for differing larval behavior PACKET MODELING Figure 6: Connectivity matrices for larval dispersal examining the role of the duration of the observation time. (a) Connectivity matrices obtained from the simulations for 1, 6 and 12 years of observation time. (b) Prediction by a simple advection-diffusion approach. (c) Prediction by the settlement-pulse model for 1, 6, 12 and 120 years of observation time. We measure spatial heterogeneity in connectivity by using the coefficient of variation of the connectivity along the mean source location line (black dash-dotted lines). The obtained results are 0.5, 0.25, 0.15 and 0 (upper panels from left to right) and 0.4, 0.22, 0.16 and 0.08 (lower panels from left to right) Figure 3: Connectivity matrices for larval dispersal for (a) a first realization of simulation, (b) diffusion model and c) a second realization of simulation. Source locations (vertical axis) and destination locations (horizontal axis) are identified with their alongshore location. The dashed slanted line indicates self settlement, i.e., where source locations are identical to destination locations. The connectivity matrices normalized so that the summed value becomes unity. Proposed model can capture the stochastic nature of connectivity fairly well Figure 4: Connectivity matrices examining the role of life histories. Larval settlement competency time window is set to a) 20 to 40 days and b) 5 to 10 days. The same flow field is used here. Figure 5: Connectivity matrices examining the role of larval behavior. a) Larvae are surface followers. b) Mimicking ontogenetic vertical migration by letting larvae change their depth from the surface to 30 m (after 10 days at surface). Here, the same flow field is used.