Forecasting the Spatial Structure of the Atlantic Butterfish Stock Using Survey Based Spatial Indicators, Min/Max Autocorrelation Factor Analysis and Kriging

The spatial structure of the Atlantic butterfish stock was investigated using survey based spatial indicators and min/max autocorrelation factor (MAF) analysis. In addition to the standard approach of combining indicators into MAFs, indicators that carried the major signals were selected and used to forecast the future status of the stock by kriging the MAFs. Northeast Fisheries Science Center fall survey data, 1982–2013, were used to characterize the population location (center of gravity) and occupation of space (inertia, positive area, spreading area, and equivalent area) over the time series. The three highest loadings on MAF1 were the latitude of the center of gravity for age 2+ butterfish, followed by the positive area for age 0 and age 2+ butterfish, respectively. A 6-year forecast for MAF1 beginning in 2013 was done using the historical data. Although the kriging error and confidence interval increased with time, the predicted 2013 MAF was nearly coincident with the actual MAF based on observed data for 2013. The short term prediction is that the center of gravity for age 2+ butterfish would move south, the positive area for age 0 fish would increase, while the positive area for age 2+ butterfish would decrease.