A photogrammetric methodology for size measurements: application to the study of weight–shell diameter relationship in juvenile Cantareus aspersus snails

Abstract

A photogrammetric methodology is proposed to measure the diameter of snail shells on digital pictures. Digital photographs were taken of juvenile Cantereus aspersus snails. AutoCAD image analysis software was used for the measurements and the results were contrasted with data obtained using a digital caliper in the same snails to compare the accuracy of both methods. The snails were individually weighed and the shell diameter was measured once a week, for a total of 7 weeks. After the third week, there were no significant differences between both methods, whereas in the first 2 weeks the measurements obtained with a digital caliper scored larger diameters than the photogrammetric measurements. This is probably due to the difficulty to define the end of the shell with the caliper, whereas the photogrammetric analysis does not involve any risk for the snail. To test the reproducibility and repeatability of both methodologies seven snails were measured five times by three different examiners. Using the variance components analysis, the repeatability was 4.8% of the total variation for the photogrammetric methodology and 6.9% for the conventional methodology, while the reproducibility was 0.0% and 2.7% for the photogrammetric and conventional methodology, respectively. These findings indicate that the photogrammetric methodology can be used with confidence to measure the diameter of snail shells. The great advantage of this method is the ability to magnify the images to make more precise measurements from snails of any size, which is especially useful in the early stages of development. The growth data were used to construct a model for live weight (LW) estimation based on shell diameter. The snails showed high growth rates both in terms of shell diameter and LW. The shell diameters showed a low individual variation (16.2% CV) and were normally distributed, whereas the LWs showed greater variability (36.8% CV) and were not normally distributed. The best model for estimating LW from shell diameter in juvenile snails is the multiplicative model, with a high determination coefficient (96%).