Thirteen Years of Monitoring a Local Population of Eritrichium caucasicum: Stochastic Growth Rate under Reproductive Uncertainty

• Published in: Biology bulletin reviews Vol. 14; no. 1; pp. 73 - 84
• Main Authors: Logofet, D. O., Golubyatnikov, L. L., Kazantseva, E. S., Ulanova, N. G., Khomutovsky, M. I., Tekeev, D. K.
• Format: Journal Article
• Language: English
• Published: Moscow Pleiades Publishing 2024; Springer Nature B.V
• Subjects: Biochemistry; Biomedical and Life Sciences; Cell Biology; Ecology; Endemic species; Eritrichium; Flowering; Growth rate; Life Sciences; Population dynamics; Stochasticity; Weather; Zoology; Caucasus
• ISSN: 2079-0864; 2079-0872
• DOI: 10.1134/S2079086424010055

— Eritrichium caucasicum is an alpine short-lived perennial species endemic to the Caucasus. The stage structure of a local population has been observed on permanent plots in the alpine belt of the Northwestern Caucasus annually for 13 years (2009–2021), accumulating data of the “identified individuals from unknown parents” type. The latter circumstance has predetermined what is called reproductive uncertainty in the terminology of matrix models for discrete-structured population dynamics and means that the annual recruitment rates inherent in the groups of generative plants and final-flowering generative plants cannot be calibrated in a unique way. As a result, instead of the annual values of the asymptotic growth rate, the model gives only certain ranges of their values that vary from year to year, corresponding to the data. This introduces both technical difficulties and uncertainty in the viability forecast based on the asymptotic growth rates. A well-known alternative approach is to estimate the stochastic growth rate λ S , but only artificial modes of randomness involved in the calculation of λ S have been proposed in the literature. Our realistic model of randomness is related to variations in weather and microclimatic conditions of the habitat. It is reconstructed from a fairly long (60 years) time series of the weather indicator. Using this realistic model in Monte Carlo calculations of λ S , we have obtained a more reliable and accurate estimate of the stochastic growth rate.