Neural network prediction of biodiversity and biometric indicators in herring (Clupea pallasii) of Okhotsk Sea
Abstract
The paper presents data on the prediction of morphometric measurements for herring of the Sea of Okhotsk using machine learning algorithms (neural networks). It has been shown that at least 12 features correlate with each other with a high and very high degree of correlation (r>0.7–0.9). The possibility of using neural networks to predict missing morphometric (and other types of) data for any biological objects, regardless of their geographical habitat, is demonstrated. As an example, the values of 6 morphometric features were given, such as: the length of the entire fish (ab), the weight of the fish, the distance between P and V (vz), the lengths of the lower and upper blades C, the anteanal distance. The predicted values, as shown, deviated from the reference values for a number of measurements from 0.2% to 3%, which is less than the natural variance in the sample, reaching up to 14% in some respects. Everything presented allows us to propose neural networks as a modern scientific method, for example, to eliminate the lack of statistical data or to «close the needs» for obtaining new morphometric measurements.
About the Authors
V. V. Gorbachev
Plekhanov Russian University of Economics; Magadan Branch of VNIRO Federal State Budgetary Budgetary Institution (MagadanNIRO)
Russian Federation
Russian Federation
Viktor V. Gorbachev – researcher at the Research Laboratory «Biotechnology of Food Systems» of the Department of Food Technology and Bioengineering
115054, Moscow, Stremyanny Lane, 36
A. A. Smirnov
All-Russian Scientific Research Institute of Fisheries and Oceanography (VNIRO); Northeastern State University (SVSU); Dagestan State University (DSU)
Russian Federation
Russian Federation
Andrey A. Smirnov – Doctor of Biological Sciences, Associate Professor, Chief Researcher of the Department of Marine Fishes of the Far East; Professor of the Department of Exact and Natural Sciences; Associate Professor of the Department of Ichthyology
105187, Moscow, Okruzhnoy proezd, 19
685000, Magadan, Portovaya str., 13
367025, Makhachkala, Gadzhieva str., 43a
E. A. Metelyov
Federal State Budgetary Scientific Institution VNIRO (MagadanNIRO)
Russian Federation
Russian Federation
Evgeniy A. Metelyov – Candidate of Biological Sciences, Head of the Magadan Branch
685000, Magadan, Portovaya str., 36/10
References
1. Convention on Biological Diversity of June 5, 1992. // “Collection of legislation of the Russian Federation. Moscow. 05/6/1996. p. 2254.
2. Rogers A.D., Appeltans W., Assis J. [et al.] (2022). The discovery of marine biodiversity in the 21st century // Adv. Marine Biology. Vol. 93. Pp. 23-115. DOI:10.1016/bs.amb.2022.09.002
3. Bef G. (2011). Characteristics of marine biodiversity. // C.R. Biol. V. 334. Pp. 435-440. DOI:10.1016/j.crvi.2011.02.009. Biological Journal
4. Miller R.S., Odum E.P. Fundamentals of ecology // Oikos. 1954. 134 p. DOI:10.2307/3564656
5. Ney M., Kumar S. Molecular evolution and phylogenetics – Oxford: Oxford University. Click. 2000. 333 p. ISBN 978-0-19-513585-5
6. Gorbachev V.V. Migration as a cause of genetic homogeneity of Pacific herring (Clupea Pallasii) from the Sea of Okhotsk // Russian Journal of Applied Genetic Research. 2013. Vol. 3. pp. 203-208. DOI:10.1134/S2079059713030052
7. Gorbachev V.V., Lapinsky A.G., Prikoki O.V., Solovenchuk L.L. Modeling the dynamics of the effective pollock population of the Sea of Okhotsk in the Holocene epoch based on genetic variability in mtDNA Nd2 and Cytb loci // Russian Journal of Genetics. 2014. vol. 50. Pp. 763-768. DOI:10.1134/S1022795414070072
8. Smirnov A.A. Biology, distribution and condition of stocks of Gizhiginsky-Kamchatka Herring. – Magadan: SVSU. 2014. 179 p. (In Rus.)
9. Nikolenko S., Kadurin A., Arkhangelskaya E. (2018). Deep learning. Introduction to the world of neural networks – St. Petersburg: Peter. 480 p. (In Rus.)
10. Samuel O.M., Casanova P.M., Olopade J.O. (2018). Elliptical Fourier contour descriptors and morphological analysis when looking from behind at the occipital foramen of a tropical raccoon (Procyon Cancrivorus) (Linnaeus, 1758) // Morphology. Issue 102. Pp. 31-40. DOI:10.1016/j.morpho.2017.06.001
11. Nurdalila A.A., Bunawan H., Kumar S.V. [et al.]. (2015). The homogeneous nature of the Malaysian marine fish Epinephelus Fuscoguttatus (Perciformes; Serranidae): evidence based on molecular markers, morphology and infrared analysis with Fourier transform // Int J Mol Sci. V. 16. pp. 14884-14900. DOI:10.3390/ijms160714884
12. Bahmani L., Abunajmi M., Arabhosseini A., Mirsaedgazi H. (2018). Modeling of the kinetics of extraction of essential oil from tarragon using ultrasound pretreatment // Engineering in agriculture, environment and food. Vol. 11. Pp. 25-29. DOI:10.1016/J.eaef.2017.10.003
13. Gorbachev V.V., Nikitina M., Velina D. [et al.]. (2022). Artificial neural networks for predicting the antiradical potential of food products // Applied Sciences. Vol. 12. p. 6290. DOI:10.3390/app12126290
14. Behruzi K. Tavakoli T., Ghassemyan H., Khoshtagaza M.H., Banakar A. (2013). Application of machine vision and artificial neural network for modeling and controlling the process of drying grapes // Computers and electronics in agriculture. Vol. 98. Pp. 205-213. DOI:10.1016/j.compag.2013.08.010
15. Sokal R.R., Rolf F.J. (2012). Biometrics: principles and practice of statistics in biological research / R.R. Sokal, [ed.] 4th ed. – New York: Freeman W.H. 915 p. ISBN 978-0-7167-8604-7
Review
For citations:
Gorbachev V.V., Smirnov A.A., Metelyov E.A. Neural network prediction of biodiversity and biometric indicators in herring (Clupea pallasii) of Okhotsk Sea. Fisheries. 2024;(3):32-39. (In Russ.)
Views:
24
Email this article 