Mark–recapture versus length–frequency based methods: Evaluation using a marine gastropod as a model

• The growth estimates calculated by mark–recapture and length–frequency data were compared.
• The two methods applied for each method only presented differences at the 1st year.
• A new method blending the two was proposed to soften the differences.
• The credibility of indirect method is important for certain groups like invertebrates.

This work aims to compare growth estimates calculated by two types of data, obtained directly (mark–recapture) and indirectly (modal progression analysis based on monthly length–frequency data of a population) using a Tegula viridula marine snail population from Flexeiras Beach, Itacuruçá Island, Rio de Janeiro as model. During the study period (May/2010–May/2012), 3789 snails were collected and measured; among these, 765 were tagged and 362 were recaptured, some more than once. The population parameters of the seasonalized von Bertalanffy growth function estimated by the two methods presented differences in the first year (Direct: L∞ = 31.23 mm, K = 0.49 yr− 1, C = 0.20, Wp = November; Indirect: L∞ = 30.60 mm, K = 0.65 yr− 1, C = 0.29, Wp = May; ARSS: F3,64 = 20.90, p < 0.05), which, were not significantly different in the second year (Direct: L∞ = 27.40 mm, K = 0.85 yr− 1, C = 0.30, Wp = March; Indirect: L∞ = 27.29 mm, K = 1.00 yr− 1, C = 0.37, Wp = March; ARSS: F3,48 = 1.98, P > 0.05). The differences between the methods in the first year seem to be related to the sample size effect, especially regarding the direct method (first year: n = 142, second year: n = 773). Validating this statement, when the data from both years were combined, the differences between the methods were softened (Direct: L∞ = 27.44 mm, K = 0.84 yr− 1, C = 0.30, Wp = March; Indirect: L∞ = 28.84 mm, K = 0.83 yr− 1, C = 0.25, Wp = March; ARSS: F3,123 = 1.46, P > 0.05). Mark–recapture analyses are usually considered to providing better growth estimates than indirect methods, and are also assumed to be less dependent on sample sizes. A new methodology is proposed, combining the two methods, using the length of recaptured individuals as cohorts and applying a modal progression analysis.