Question:

Where are the statistically significant regions of high and low Alaska Skate catch rate from commercial longliners in the Bering Sea?

Name of the tool or approach used.

ArcGIS- Hot Spot Analysis

Brief description of steps you followed to complete the analysis.

The dataset that I used to complete this analysis includes over 200 species of fish and four gear types, so I chose to limit the scope of this exercise to one species in one gear type.  So I narrowed the dataset by selecting Alaska Skates caught by longliners, and ran a Hot Spot Analysis on this selection.  Within the Hot Spot tool in Arc, I selected catch “RATE” as the input field for comparison, retained the default selections for the spatial relationship (“FIXED_DISTANCE_BAND) and distance method (“EUCLIDEAN_DISTANCE”), and ran the analysis.  
Narrowing the temporal scope of the analysis, I ran it again on two subsets of the data.  One on points from 1998-2007 and the second on points from 2008-2014.

Brief description of results obtained.

The original analysis includes AK skate catch rates on Bering Sea longliners from 1998-2014 and reveals some interesting patterns about fishing behavior as well as where skate catch rates are higher.  There is a clear line of cold spots along the southern edge of the oceanic shelf that runs north of the Aleutian Islands.  A few distinct hotspots are also revealed in the western Aleutians near Adak and Atka islands

Changing the temporal scope of the analysis revealed some very different patterns than the full time period.  There are a number of hypotheses that can be developed for further exploration as to why the patterns change.  

Critique of the method – what was useful, what was not?

For this dataset, I think that this tool is a quick and easy way to identify patterns at different scales and could be an approach to hypothesis generation.  This dataset is one that the sample points are independent of the researcher and span a large temporal and spatial range.  It seems that for this tool these characteristics in the dataset make it more applicable.