Errors in juvenile copepod growth rate estimates are widespread: Problems with the Moult Rate method

Abstract

The 'Moult Rate' (MR) method has been used widely to derive stage-specific growth rates in juvenile copepods. It is the most common field-based method. Unfortunately, the equation underlying the method is wrong and, consequently, large errors in juvenile growth rate estimates are widespread. The equation derives growth from the mean weight of 2 consecutive stages (i and i + 1) and the duration of stage i. The weight change and the period to which this change is attributed are, therefore, offset. We explore this potential source of error in the MR method critically. Errors arise as a result of 2 primary factors: (1) unequal durations of successive stages and (2) unequal rates of growth of successive stages. The method of deriving the mean weight (arithmetic or geometric) also has an impact and is examined. Using a steady-state assumption, a range of scenarios and the errors that arise are examined. The literature is then reviewed and the size of errors resulting from MR method application in both field and laboratory situations is estimated. Our results suggest that the MR method can lead to large errors in growth estimation in any stage, but some stages are particularly prone. Errors for the C5 stage are often large because the following stage (the adult) does not moult, and has a different rate of body weight increase. For the same reason, errors are also great where the following stage is not actively moulting (e.g. when diapausing). In these circumstances, published work has commonly greatly underestimated growth. For example, MR growth ranges from 11 to 47% of the value derived correctly for this stage, gi_corr (calculated assuming the non-moulting stage does not grow). In late stages that are followed by actively moulting stages, the MR method has commonly given values in excess of 150% of gi_corr, but underestimation also occurs, with values <90% of gi_corr. We propose new methods and equations that overcome these problems. These equations are written with and without within-stage mortality included. The equations are relatively insensitive to mortality rates within the range found in the field, but only provided that the stage duration is not determined from moult rate. Stage duration estimates obtained from measuring moulting rates of field-collected animals are very sensitive to mortality rates of the animals prior to capture, and field mortality rates are often high enough to produce dramatic over-estimation of stage duration. © Inter-Research 2005.