Most sessile marine animals have a larval stage that can potentially disperse long distances.  Consequently, animals harvested in one spot may have been born in a completely different location, and understanding how and where they moved is critical for managing harvested species and deciding how to position marine reserves.  Our project evaluated larval dispersal in commercially-harvested blue mussels in eastern Maine, where a strong coastal current is located offshore of a very complex coastline, with deep bays separated by rocky headlands and islands.  Because mussels live up in the bays, the dispersal of mussel larvae up and down the bays and in and out of the coastal current plays an important role in determining dispersal patterns.  We used a coupled biological/physical model to predict mussel dispersal patterns.  This model was based on a very high-resolution coastal circulation model (the physical component) which we tested with field-collected current data.  The model was very accurate at predicting observed currents.  This physical model was then coupled with a biological model of larval mussel growth, mortality, and behavior.  The growth and mortality parts of the model were based on laboratory experiments in which we reared larvae at different temperatures, while the behavior component was based on field observation of the vertical movement of mussel larvae.  Contrary to some past reports, we discovered that mussel larvae migrate vertically on a day/night cycle, but do not migrate vertically on the flood/ebb tidal cycle.  They do move vertically on the tide cycle, but this movement is passive and driven by currents, and thus does not represent a behavior that needs to be modeled.

The coupled biophysical model predicted that most larval dispersal occurs among mussel beds within a bay, or between neighboring bays.  Successful dispersal over longer distances did occur, but much more rarely.  Dispersal was mostly from northeast to southwest (in the direction of the coastal current), but localized northeastward dispersal against the prevailing current was possible because flooding tides can move the larvae between bays within one tidal cycle.  Tides were the main process moving larvae between the bays and the coastal current, while the physical structure of the bottom largely dictated where these across-shelf movements occurred.  Wind appeared to play surprisingly little role in moving larvae up and down bays, but that result is partly because wind direction did not vary much.  This region is known as "downeast" Maine because the wind blows very consistently from the southwest.

We engaged with mussel harvesters and managers (the Maine state Dept. of Marine Resources) to share data and results.  Mussel harvesting has traditionally been largely unregulated in Maine, but market pressures driven by the need to document that the harvest is sustainable are pushing the industry toward greater regulation.  The state has only recently begun monitoring this resource, and their efforts to date have been limited to one bed.  We were able to supply preliminary data on 14 other beds over a 4 year time span.  Those data had originally been collected to support our modeling efforts (we needed to know how many larvae each bed produced, which is determined by the number of mussels and how large they are), but served as a preliminary stock assessment. 

We also trained two post-docs and eight undergraduate research assistants.  One of the undergraduates worked on the project two summers and developed an independent project that was recently published in a major journal; she is currently applying to PhD program.  One post-doc now has a permanent job in industry, and the other has a Knauss fellowship from NOAA to work with the federal legislature for a year.