Morphological Identification of Bighead Carp, Silver Carp, and Grass Carp Eggs Using Random Forests Machine Learning Classification

Abstract

Visual identification of fish eggs is difficult and unreliable due to a lack of information on morphological egg characteristics of many species. We used random forests machine learning to predict the identity of genetically identified Bighead Carp (Hypophthalmichthys nobilis), Grass Carp (Ctenopharyngodon idella), and Silver Carp (H. molitrix) eggs based on egg morphometric and environmental characteristics. Family, genus, and species taxonomic level random forests were explored to assess variable performance and accuracy. Bighead Carp, Grass Carp, and Silver Carp egg characteristics were similar and difficult to distinguish from one another. When combined into a single invasive carp class, random forests were ≥97% accurate at identifying invasive carp eggs with a ≤5% false positive rate. Egg membrane diameter was the most important predictive variable, but the addition of ten other variables resulted in a 98% success rate for identifying invasive carp eggs from 26 other upper Mississippi River basin species. Our results reveal that a combination of morphometric and environmental measurements can be used to identify invasive carp eggs. Similar machine learning approaches could be used to identify eggs of other fishes. These results will help managers more easily and quickly assess invasive carp reproduction. Disciplines Disciplines Aquaculture and Fisheries | Natural Resources Management and Policy | Statistical Models Comments Comments This is a manuscript of an article published as Camacho, Carlos A.,

Evaluation of a Random Forest Model to Identify Invasive Carp Eggs Based on Morphometric Features

“…Sampling of fish eggs in this study was performed in the same manner as described by Camacho et al. (2019) and was conducted every 10 d from late April through September at a series of sites at tributary confluences to the UMR or within a specific tributary reach. Confluence sites were composed of a total of three transects: (1) within the tributary, (2) within the UMR approximately 1 km upstream of the tributary mouth, and (3) within the UMR approximately 1 km downstream of the tributary mouth.…”

“…We measured the same egg morphometrics and characteristics (13 biotic, 4 abiotic) that were described and identified as useful predictor variables by Camacho et al. (2019). These predictor variables were obtained from the pictures of each egg sent for genetic analysis and were used with the model to aid in classifying species by their unique traits (Table 1).…”

“…Abiotic variables included water temperature (°C), conductivity (μS/cm), Julian day, and month of egg collection to account for variable spawning condition requirements (Camacho et al. 2019, 2021).…”

“…A recent study demonstrated the efficacy of morphological identification of invasive carp eggs by using random forest machine learning models produced from genetically identified eggs (Camacho et al. 2019). Random forest models are powerful predictive models with high accuracy (Breiman 2001).…”

“…The objective of this study was to validate the Camacho et al. (2019) set of models for identifying invasive carp eggs from morphological characteristics. First, we reproduced the models from Camacho et al.…”