Populations of most marine organisms are connected by the dispersal of larval stages, with profound implications for marine conservation. Because of the extreme effort needed to empirically measure larval exchange, multispecies conservation efforts must estimate connectivity by extrapolation using taxonomy, adult distribution, life history, behavior, or phenology. Using a 6-year record of connectivity realized through trace-elemental fingerprinting of larval shells, we document the seasonal and interannual variability of larval exchange for two congeneric mussel species with overlapping but distinct distribution, life history, and reproduction timing. We reveal consistent autumn poleward movement and spring equatorward movement for both species, coincident with near-shore surface currents. However, because the major reproductive seasons differ, the dominant source-sink dynamics of these two congeneric species are nearly opposite. Consideration of present and future reproductive timing as altered by climate change is crucial to marine connectivity and conservation, especially for the numerous coastal areas subject to seasonal current reversals.

The design of marine reserves is complex and fraught with uncertainty. However, protection of critical habitat is of paramount importance for reserve design. We present a case study as an example of a reserve design based on fine-scale habitats, the affinities of exploited species to these habitats, adult mobility, and the physical forcing affecting the dynamics of the habitats. These factors and their interaction are integrated in an algorithm that determines the optimal size and location of a marine reserve for a set of 20 exploited species within five different habitats inside a large kelp forest in southern California. The result is a reserve that encompasses approximately 42% of the kelp forest. Our approach differs fundamentally from many other marine reserve siting methods in which goals of area, diversity, or biomass are targeted a priori. Rather, our method was developed to determine how large a reserve must be within a specific area to protect a self-sustaining assemblage of exploited species. The algorithm is applicable across different ecosystems, spatial scales, and for any number of species. The result is a reserve in which habitat value is optimized for a predetermined set of exploited species against the area left open to exploitation. The importance of fine-scale habitat definitions for the exploited species off La Jolla is exemplified by the spatial pattern of habitats and the stability of these habitats within the kelp forest, both of which appear to be determined by ocean microclimate.

[1] Observations from autumn 2000 near the shelfbreak front in the Middle Atlantic Bight are used to describe the transition from stratified summer conditions to well-mixed winter conditions over the shelf. During the observational period, the front differed dramatically from climatological conditions, with buoyant Gulf Stream water found shoreward over the subsurface shelfbreak front. Water mass analysis shows a large number of separate water masses with shelf, slope, and Gulf Stream origins. The coolest shelf water was located at the shelfbreak and may be related to ''cold pool'' water masses observed to the north during summer. Shoreward of this shelfbreak water mass, a mid-shelf front was present which intersected the bottom at the 50-m isobath. High-volume transports were associated with both the shelfbreak and mid-shelf fronts. Transport estimates from the cross-shelf sections were approximately 1 Sverdrup, which is large relative to previous estimates of shelf transport. The foot of the front was near the 130-m isobath, much deeper than the climatological position near the 75-m isobath; however, this is consistent with a recent theory relating the magnitude of alongshelf transport to the depth at which the front intersects the bottom.