FISHERIES OCEANOGRAPHY COORDINATED INVESTIGATIONS (FOCI):
Responses of Ichthyoplankton Biodiversity and Dynamics to Environmental Change
Cycles between warm and cold regimes occur in the Pacific Ocean on a multidecadal
time scale. Those regime shifts can be followed by extensive changes in
the aquatic community. We are analyzing the biodiversity and dynamics
of ichthyoplankton assemblages in more than 24 years of data collected
from the Gulf of Alaska (GOA) assembled by the AFSC Recruitment Processes
Program. The goal of our study is to create a variety of sensitive indices
to be used in assessing ecosystem integrity and predicting ecosystem shifts
due to climate change.
We standardized abundance data for each of 77 ichthyoplankton taxa and
clustered them into 18 groups with the Bray-Curtis distance measure and
Flexible Beta linkage method. We analyzed the data with a variance partitioning
analysis to distinguish between geographical, seasonal, and annual variation.
Figure 4. Variance Partitioning Analysis of functional groups in the Gulf of Alaska
to distinguish between annual, seasonal, and geographical variance. Functional Groups
were named after dominating species within that group.
For many clusters most variance was explained by geographical region and
month in which the samples where taken, while the annual variation only
accounted up to 10% (Fig. 4 above). Therefore, consequent analyses focused on
the May ichthyoplankton assemblage and on each of seven geographical strata
within the GOA individually. Response variables (abundance of species
and cluster, diversity, survival index) were then linked to environmental
explanatory variables (Pacific Decadal Oscillation index (PDO), temperature,
freshwater runoff, total flow and wind indices) by canonical correspondence
and canonical correlation analysis. Temperature-related indices seemed
to be most closely linked to the response variables and diversity and survival
indices were more suitable than abundance indices alone.
In the future, we will strive to achieve improved predictions on community
change with dynamic linear modeling and Ecosim.
By Jeff Napp.
quarterly Oct-Dec 2003 sidebar
Auke Bay Lab