3. Results
3.1. Effect of the different sampling strategies
Similar outcomes were found for the single grab sampling strategy (Fig. 3), the 24-h sampling strategy (Fig. 4), the 48-h sampling strategy (Fig. 5) and the 7-d sampling strategy (Fig. 6). Except for the single grab sampling strategy in dataset 1, the chance of obtaining an estimated EF that is within 15% of the true EF was higher than the chance of obtaining an estimated EF that is more than 15% higher than the true EF (Table 1). Except for the single grab sampling strategy in dataset 2, these sampling strategies had a higher chance of estimating an EF that is more than 15% higher than the true EF, than estimating an EF that is more than 15% lower than the true EF. For the long-term weekly grab sampling strategy, the chance of obtaining an estimated EF that is within 15% of the true EF was 100% for both datasets and for all days (data not shown).