requirements, but costs $16,700 to purchase. The Satake Pulse Counter has not yet been priced 
for market, but can be used with minimal training to analyze the > 50 pm and 10-50 pm size 
classes; active processing time for both size classes is less than 5 minutes with results available 
for the > 50 pm class in approximately 10 minutes and 30 minutes for the 10-50 pm class. The 
SGS ATP (aquatools) kit requires minimal training and processes samples for either the > 50 pm 
or 10-50 pm size class in approximately 50 minutes; the initial start-up costs are $8,500 to 
purchase equipment and running costs are $35 per sample; some costs may decrease as 
devices mature from developmental technology to commercially-available units. 
4.1 Caveats and Future Directions 
In reviewing our results, it is important to keep in mind that all samples are subject to 
sampling effects and, due to random chance, it is not expected that every subsample or 
replicate measurement would contain identical numbers of organisms (Frazier et al., 2013; i.e. 
if a 10 L ballast water sample containing 1000 organisms was split into ten 1 L subsamples, we 
would not expect that each subsample would contain exactly 100 organisms). Further, the 
variation between subsamples should increase with the mean (i.e. there is more variation 
between subsamples if the true concentration is 1000 individuals / Lthan 100 individuals / L). In 
fact, variation among subsamples is expected to equal the mean if all organisms have identical 
and independent chances of being sampled, but variation could be amplified by organism 
behaviour and/or the presence of colonies in samples. Therefore, even if all methods gave 
completely accurate results, variation in measurements between subsamples is expected 
(Miller et al., 2011). Importantly, as a result, observed differences between measurements may