Heritable variation in tree growth and needle vegetation indices of slash pine (Pinus elliottii) using unmanned aerial vehicles (UAVs)

Abstract

Vegetation indices (VIs) are mainly associated with plant yield, and yield-related physiological traits can assist in breeding selection. Tree breeding programs require a rapid assessment of a large number of individual trees across multiple regions. Unmanned aerial vehicle (UAV) multispectral platforms combined with high spatial, temporal and spectral resolutions can readily assess the VIs at various plant growth phases, which has been deeply detailed for evaluating the physiological status of agricultural crops and forest stands. However, the underlying genetics of foliar spectral bands, VIs and growth traits from a large scale of samples, especially in the case of forest trees, have been less investigated. In this paper, two sites of slash pine breeding populations were completely investigated to monitor the spectral bands, VIs, and tree growth traits. This allows us to determine the genotypic variation of these traits and to estimate the heritability (h2). The results show that the estimated h2 ranges from 0.04 to 0.36, and the site may affect tree growth because the traits between two sites come to different h2 values. At the two sites, the h2 values of the green red difference vegetation index (GNDVI) and anthocyanin reflectance index (ARI) at Site 1 and the red edge chlorophyll index (RECI), leaf chlorophyll index (LCI), red green blue vegetation index (RGBVI), normalized difference red edge index (NDRE), NIR and red edge at Site 2 were less than 0.1. However, breeding selection is still possible for most VIs and growth traits at both sites based on different breeding targets and families (families 3, 5, 10, 11, 14 and 17 at Site 1 and 4, 6, 8, 19 and 13 at Site 2 for optimal spectral bands; families 2, 8, 7, 13, 16, and 18 at Site 1 and 6, 7, 11, 14, 15, 18, and 20 at Site 2 for optimal VIs; families 8 and 16 at Site 1 and families 7 and 18 at Site 2 for optimal growth traits) with a variety of optimal traits are selected. This paper demonstrates the potential use of UAV-based imagery technologies in the breeding selection of slash pine trees.