Robust Increases of Land Equivalent Ratio with Temporal Niche Differentiation: A Meta-Quantile Regression

Intercropping has been shown to be land use efficient, but there is a large variation in the land equivalent ratio (LER) among studies. We used quantile regression to estimate the effect of temporal niche differentiation and its interaction with other key characteristics of intercropping, i.e., crop type combination, N fertilizer, relative density, and intercropping pattern, using data from the intercropping literature. Quantile regression characterizes the entire distribution of the response metric by estimating quantiles of this distribution. This method gives a comprehensive characterization of the diversity of the response in the population. In this study, the effect of temporal niche differentiation on the LER was positive and significant across all quantiles (10, 15, ..., 85, 90%). The response of the LER to temporal niche differentiation was similar in C-3-C-3 and C-3-C-4 intercrops except at low quantiles (below 20%), where a stronger response was found in C-3-C-4 than in C-3-C-3 intercrops. There was a negative effect of N fertilizer on the LER in the absence of temporal niche differentiation at all tested quantiles. At low LER (quantiles <20%), this negative effect was alleviated by increasing temporal niche differentiation due to a positive interaction between N fertilizer and temporal niche differentiation. The results indicate that temporal niche differentiation has a robust positive effect on the LER across a wide range of LERs, but the strength of this effect is influenced by intercrop characteristics, especially at low to median LER values. Case-specific experiments are therefore indispensable for identifying advantageous species mixtures and optimal management adapted to local conditions.