In a recent account regarding the destruction of chloropigments within the guts of copepods, Head & Harris (1996) (H&H) presented valuable data on pigment destruction in copepods. However, in one of their main conclusions, the authors invoked 2 enzyme pools to explain the pattern of pigment destruction: one directly derived from copepods, the other one produced by the ingested algae. If this conclusion is correct, it would have tremendous iinpact on the interpretation of data collected by the gut pigment technique. Estimating ingestion rates of copepods in the field would be very difficult, if not impossible, if pigment destruction was dependent upon an unknown food composition in the gut. We therefore felt it necessary to examine the evidence presented in H&H carefully. As we will demonstrate, (1) there is no evidence to postulate the existence of 2 enzyme pools, and (2) the majority of enzymes responsible for pigment destruction are a s likely to originate from copepods a s from the ingested algae. We scanned Fig. 4b of H&H with a HP Scanjet IIC color scanner and digitized the data with the image analysis software 'Data Thief' (by Kees Huyser and Jan van der Laan, National Institute for Nuclear Physics and High Energy Physics, Amsterdam, The Netherlands). We did not directly use the data shown in Fig. 7 of H&H because the data in Fig. 4 had fewer hidden points and could therefore be reproduced more accurately. Fig. 7 was constructed by H&H by multiplying the proportion of pigment destruction by the ingestion rate. Of 102 data points used in Fig. 4 of H&H, only 80 points were visible. However, when calculating the linear regression through these 80 data points, the estimated parameters as well as the calculated coefficient of determination were very similar to the original (Fig. l a ) . H&H estimated enzyme activity by multiplying chlorophyll destruction (y-axis) by ingestion rate (xaxis), and then plotted enzyme activity versus ingestion rate creating an autocorrelation in Fig. 7 . Instead of using the percentage value, H&H multiplied the xaxis from Fig. 4 by the proportion of chlorophyll destruction ranging from ca 0.45 to l . Hence, the units of the Xand y-axes of Fig. 7 are identical (ng mg h-'). Due to this autocorrelation, one would expect a linear relationship to result from any random distribution of a y-variate multiplied by an X-variate and plotted against the X-variate. However, as is already apparent from Fig. 4 , the relationship is not exactly linear, but slightly curved. The curve begins with a slope of ca 1 at low ingestion rates (i.e. 100% chlorophyll destruction) and decreases to ca 0.45 (i.e. 45 % chlorophyll destruction) since the proportion of pigment destruction is decreasing with increasing ingestion rates. Another consequence of the autocorrelated data set is that there cannot be a n intercept, since any product of 0 equals 0. Since ingestion rate is multiplied by chlorophyll destruction to calculate enzyme activity, the product is 0 whenever ingestion rate equals 0. Instead, the positive intercept in H&H is a n artifact caused by data points in the upper part of the curve which 'drive' the linear regression, a common problem in regression analysis (i.e. outliers have more influence on the regression line and rotate the line around the mean). It is therefore no coincidence that the slope of the linear regression (Fig. 7) and the proportion of enzyme destruction at high ingestion rates (Fig. 4 ) are almost identical (i.e. approximately 0.45). In using a linear regression for displaying enzyme kinetics, H&H chose an unconventional approach. In the simplest enzyme-substrate relationship, one would expect the kinetics to follow the classic MichaelisMenten model (Michaelis & Menten 1913) and not a straight line. In this model of regulation of enzyme
Search from 250 M+ research papersSearch from 250 M+ research papers
Open Access
Mar 20, 2019
Dinesh Rano + 1
