Modeling of filter-feeding behavior in the brown mussel, Perna perna (L.), exposed to natural variations of seston availability in Santa Catarina, Brazil
The aim of this study is to quantify and model the filter-feeding behavior of the mussel Perna perna feeding on natural seston. Models were generated that described each step of the feeding process and produced a predictive model of rates of food uptake by P. perna in culture areas from Southern Brazil. Feeding experiments using the biodeposition approach were conducted with mussels ranging in shell height from 3.94 to 9.22 cm of three sites, including turbid and clear water environments. Organic content of the seston (OCS, fraction) decreased as total particulate matter (TPM, mg L-1) increased. The maximum filtration rate (FR, mg L-1) measured for an individual mussel was 156.7 mg h-1 and was recorded when TPM was 33.9 mg L-1 and OCS was 0.18. Rejection rate of particles had a strong positive relationship with TPM, and an inverse relationship with OCS. Maximum rejection rate recorded was 124.1 mg h-1 and was measured under the same seston conditions as maximum filtration rate. Net organic selection efficiency by mussels (NOSE, fraction) was related to the amount of particulate organic matter (POM, mg L-1) and particulate inorganic matter (PIM, mg L-1) available in the water. NOSE was positive below PIM values of 2 mg L-1, but had negative values when POM was above 3 mg L-1 and PIM between 2 and 15 mg L-1, and positive values when POM was below 3 mg L-1 and PIM above 15 mg L-1. Maximum NOSE was 1.71, when PIM was 1.02 mg L-1 and POM was 0.67 mg L-1. Organic content of ingested matter (OCI, fraction) had a positive relationship with NOSE and TPM. Maximum OCI was 1.24 and was measured when TPM was 33.9 mg L-1, OCS was 0.18, FR was 151.30 mg h-1, and NOSE was 1.30. The net absorption efficiency of ingested organics (NAEIO) increased with increasing OCI in a hyperbolic relationship. The net organic absorption rate (NOAR, mg h-1) increased with both FR and OCI. The coupling of the equations that described filter-feeding processes for P. perna in the STELLA software environment produced a robust model with relatively low complexity and specificity. The model can predict the P. perna feeding behavior in turbid or clear water and can be used with different species if the correct coefficients are used. The coupling of this feeding model with future models of energy budget, population dynamics, seston hydrodynamics, and primary production will be valuable for the evaluation of shellfish carrying capacity.