Multispectral indices and digital image analysis to compare seasonal nitrogen fertilization strategies for bermudagrass cultivars in a Mediterranean climate

Bermudagrass is widely cultivated as turfgrass throughout the warmer areas of the world. However, it is not cold-resistant and enters into dormancy when average daytime temperatures drop below 10ºC. To reduce this dormancy period, a nitrogen fertilization strategy including late-season applications, could be effective. However, this fertilization management poses a risk of nitrogen loss, which could be mitigated by using nitrogen derived from organic sources. Nevertheless, there are few reports available on the quality of turfgrass with different nitrogen sources in Mediterranean climates, where winters are mild and bermudagrass dormancy can be minimal. For this purpose, a two-year study was conducted in Valencia (Spain) to compare two different N-foliar (organic vs mineral) applied fertilization programs on five bermudagrass cultivars (‘Arden 15’, ‘Blackjack’, ‘Common’, ‘MBG 002’, and ‘Paul 1’) from October before winter dormancy to May after full spring green-up recovery. Data acquisition from aerial digital imagery, ground-reflectometry, and conventional visual turfgrass evaluation methods were compared. Visual results showed that MBG 002 performed better in autumn and late spring than all other cultivars in terms of visual quality (maximum difference of 1.8), single 670, 730 and 780 nm reflectance, and reflectometry indices during the first year. The Normalized Difference Vegetation Index, showed a maximum difference of 1.0. However, the MBG 002 cultivar had a longer dormancy period in the warm 2021–2022 winter than Blackjack and Paul 1. Secondly, organic N-fertilization based on amino acids equaled the effect of mineral fertilization. Finally, weekly reflectometry measurements were sensitive and strongly correlated to conventional visual subjective quality evaluation (Spearman correlation coefficient, ρ = 0.85–0.89), especially the reflectance at 670 and 730 nm in the red edge bandwidth region, which was able to detect differences among cultivars and between N-programs even during the dormancy period. However, turfgrass biomass, and N and chlorophyll leaf content were poorly correlated to reflectometry indices (ρ < 0.26) and visual quality (ρ < 0.20), maybe due to the shortage of sampling events. This research shows reflectometry is a useful tool for turfgrass evaluation, and that amino acid-based fertilizers are a viable alternative to mineral nitrogen fertilization.