STUDY OF THE COMPONENTS OF DEVELOPMENTAL ONTOGENETIC NOISE USING THE EXAMPLE OF TILIA CORDATA MILL

Two regions of the distribution of small-leaved linden (Tilia cordata Mill.) have been considered: the Kola Peninsula and the Moscow region. Leaf collection was carried out to determine the fluctuating asymmetry of bilaterally symmetrical traits (FA), the noise of ontogenetic development as a fraction of trait variation devoid of fluctuating asymmetry (DN), and fitness (a measure of trait correlation). Research methods included clustering techniques and neural network modeling of FA values (multilayer perceptron with automatic neural network search). Clustering allowed the separation of populations into two groups differing in FA and fitness. In five northern populations, a combination of low fitness and high FA was observed. Conversely, in Moscow populations, modeling revealed elevated FA values alongside high leaf fitness. The authors attribute the increased asymmetry, low fitness, and high level of ontogenetic development noise (p < 0.001) in the northern part of the linden’s range primarily to high variability of leaf plates due to specific climatic conditions affecting their formation. In both regions, the level of ontogenetic noise was found to be much greater in variance than the level of FA or fitness. In Moscow populations, modeling revealed a positive correlation between FA and fitness and a negative correlation between FA and developmental noise. Northern populations, characterized by greater variance in measured traits and smaller leaves, exhibited the highest developmental instability. For modeling the level of this instability, ontogenetic development noise can be used as a determined genotypic component.