Fisheries data not current?

[eschemmel]

I had a question on the fisheries data for the 2015 global assessment. Is the mean catch data and b_bmsy data current (up to 2015)? Does the catch and b_bmsy data have to have the same years represented? For example can you have catch data from 2010-2015 but use b_bmsy data from 2005-2010?

[jamiecmontgomery]

Great question.

The OHI global assessments use reconstructed catch data from the Sea Around Us Project. This catch data is only provided from 1950 through 2010. Thus we do not have global catch data current to 2015. This catch data is used to calculate B/Bmsy values for the same time period. Since this is the most current available data we will use the 2010 data in our 2016 global assessment.

The ideal dataset would have B/Bmsy values and catch values for the same years. If that isn't available you can explore a few different options.

You could look at the B/Bmsy values for each stock, and if they don't change dramatically over time it seems reasonable to assume the same B/Bmsy for current years. This is also assuming catch levels were similar in 2005-2010. Predicting future b/bmsy values with a regression model might also be reasonable.

In the global assessment, we use B/Bmsy values from the RAM legacy database wherever possible. When we use these values, they often aren't provided for every year. We gapfill this data into the future using regression models. When doing this, we check the B/Bmsy values to make sure they seemed reasonable to fit the missing data using a regression model. We also have only predicted up to 3 years of missing data.

I hope this helps.

[eschemmel]

Here is a reference table of our catch and target year ranges https://github.com/OHI-Science/cnc/blob/draft/prep/FIS/Fisheries_CSV_Layers/fish_target_table.csv. I guess I still am curious if you can use catch data that doesʻt over lap with years that have targets. The catch stays pretty consistent so I could use average catch to gap fill in the years that do not have data. Or could just use the most recent years of data to calculate scores and trends even if they do not over lap? We only have one region so there is not an issue with weighting regions by the percentage of catch. We just need a score for each species. Your suggestions are very appreciated.

[jamiecmontgomery]

I want to make sure I understand everything correctly.

For the Global OHI assessments, an estimate of B/Bmsy was calculated for some or all of the stocks included in the table. For some of these stocks, at least the first 5 in the table, you have better information on catch and status (referred to as 'value' in your table). For those first five stocks it seems there is quite a bit of overlap in catch and status, although not perfect. I would suggest using the most recent estimate for status and the most recent estimates on catch to calculate the stock specific score. This is operating under the assumption that the status has not changed since it was last assessed.

For example, the B/Bmsy status (2.86) for Thunnus alalunga can be carried forward from 2013 to 2014 and 2015.

In your data, is the target "value" a single value or is it the average across all target years? If you have values for each target year that would be a better. This would give some perspective on how constant the values are across time and if they are fairly constant, carrying the most recent target value forward to align with catch would make sense.

Additionally, you mention that your assessment only has one region therefore you don't need to weight regions by catch. This is fine, you can calculate this goal for a single region. But you will need the catch information to weight each stocks contribution to the overall goal status score for the single region. I just want to make sure that is clear.

[eschemmel]

This makes sense. Thank you very much for the advise. The value in the table is the most recent year sb/sbmsy value. There is a time series of sb/sbmsy values that we can calculate from a regression the 2014 & 2015 values. This would let us use the most recent data for catch. I understand that we need the catch information to weight the stockʻs contribution to the overall goal score. Thank you again for the clarification.