Effects of shade cultivation and deficit irrigation on the growth and canopy structure of Coffea arabica L. in dry-hot region

Highly efficient production of Coffea arabica L. is limited due to irrigation and light management challenges in dry-hot regions. To investigate the effects of an irrigation deficit on the growth and canopy structure of C. arabica under the shade of banana plants, a field experiment was performed using different irrigation levels and shade cultivation modes. A complete combination design was adopted, with three levels of irrigation, i.e. full irrigation (FI), light-deficit irrigation (DI_L) and severe deficit irrigation (DI_S), and four modes of shade cultivation, i.e. no shade cultivation (S₀:monoculture C. arabica), light shade cultivation (S_L:intercropping of four lines of C. arabica and one line of banana), moderate shade cultivation (S_M:intercropping of three lines of C. arabica and one line of banana) and severe shade cultivation (S_S:intercropping of two lines of C. arabica and one line of banana). Compared with FI treatment, other treatments of irrigation decreased the height increment and shoot length by 13.62%-23.94% and 8.82%-13.96%, respectively, and increased the total fixed-point factor by 9.55%-34.97%. Compared with S₀ treatment, other shading cultivation treatments increased the height increment, stem diameter increment, crown width increment, shoot length, and leaf area index of the canopy by 18.33%-33.65%, 6.43%-15.47%, 5.38%-12.60%, 8.82%-24.69% and 5.18%-22.85%, respectively; and decreased the opening, gap fraction, mean leaf dip angle, total fixed-point factor and transmittance of the canopy by 4.42%-15.50%, 4.85%-16.49%, 5.50%-15.07%, 13.78%-41.44% and 10.36%-31.78%, respectively. The opening, gap fraction, mean leaf dip angle, direct fixed-point factor, indirect fixed-point factor, total fixed-point factor, direct radiation of the crown, indirect radiation of the crown, transmittance, and extinction coefficient of the C. arabica canopy were significantly positively correlated with each other and were significantly negatively correlated with leaf area index. Cluster analysis indicated that the twelve treatments could be divided into three categories. The first category was FIS_S, DI_LS_S and FIS_M; the second category was FIS_L, DI_LS_L, DI_LS_M, DI_SS_S, FIS₀ and DI_SS_M; and the third category was DI_LS₀, DI_SS_L and DI_SS₀, when the distances between classes was five. It was clear that the best growth condition was the first category, which can be used as a suitable combination of irrigation and banana shade cultivation mode for C. arabica in dry-hot regions. The results of this study provide theoretical guidance for the irrigation and light management of C. arabica in dry-hot regions.