Figure 4.1. Relationship between estimated sample number and percentage change in metric for benthic data.
Figure 4.1 further illustrates this point, demonstrating that the survey effort undertaken in East Anglia THREE and FOUR can detect change well below the 50% level for each metric and Area. This is evident from the decay and stop in the line at an ESN of approximately 2 before reaching an EDPC of 50% (as it is based on the variability of data between samples, Power analysis cannot predict a sample size less than 2).
4.3.2 Cable Corridor
Table 4.4 presents the Power inherent from the 2010 surveys (MESL 2011) for benthic surveys from the cable corridor.
Table 4.4 Results from Power analysis for Cable Corridor Effect Size
2011 density (n = 71)
0.2 0.5 0.8
Power H' Power S 0.66 (99) 0.99 1
1 1 1
As with the Shannon metric from East Anglia THREE, the Power of a comparison using Shannon at an effect size of 0.2 is lower than a Power of 0.8. The 0.8 es however, equates to an estimated detection level of 18% (Table 4.2). As the 0.8 es for Simpson is also well below the 50% threshold (see Table 4.2), we can state that a survey effort of 71 stations is sufficient to characterise the site and allow a future detection of change. Figure 4.2, below, further highlights that a survey effort based on 71 sites within the cable area can detect change well below the 50% level for each metric and area.
