Remote sensing application to estimate fish kills by Saprolegniasis in a reservoir

Abstract

Saprolegniasis is one of the most economical and ecologically harmful diseases in different species of fish. Low water temperature is one of the most important factors which increases stress and creates favourable conditions for the proliferation of Saprolegniasis. Therefore, the monitoring of water surface temperature (WST) is fundamental for a better understanding of Saprolegniasis. The objective of this study was to develop a predictive algorithm to estimate the probability of fish kills caused by Saprolegniasis in Río Tercero reservoir (Argentina). WST was estimated by Landsat 7 and 8 imagery using the Single-Channel method. Logistic regression was used to relate WST estimated from 2007 to 2017 with different episodes of fish kills by Saprolegniasis registered in the reservoir during this period of time. Results showed that the algorithm created with the first quartile (25th percentile) of the WST values estimated by Landsat sensors was the most suitable model to estimate Saprolegniasis in the studied reservoir.