Sample size is small enough to make it inconclusive without using three years of data and averaging and RSB also pointed out a shift in the weighting by sample locations.
The shift to 3 year averages would not account for a 5% drop in breeding rates and neither would a shift in the weighting by sample locations. The data I posted shows quite clearly that those WMUs with the highest sample size did not decrease in productivity,so there is no reason to conclude breeding rates decreased in those WMU's
Since breeding rates were expected to increase in all WMU's the changes you and RSB noted simply don't account for a 5% decrease.