There are many indices available to evaluate the potential advantages of intercrops and species interactions but correct choice of index is crucial in making accurate interpretations. This study compared and evaluated the relevance in understanding intercrop functioning of some well-known indices (aggressivity, AG; cumulative relative efficiency index, REIc; land equivalent ratio, LER) and other potentially useful indices (change in contribution, CC; interspecific and intraspecific interaction index, IE and IA; comparative absolute growth rate, CGR). Data collected from a two-year field experiment in SW France with different fertiliser N levels comparing wheat ( Triticum turgidum L., cv. Nefer) and pea (winter pea, Pisum sativum L., cv. Lucy) grown as sole crops or intercrops in a row substitutive design were used to calculate, compare and evaluate the relevance of the selected indices for understanding intercrop functioning. It was found that AG indices (calculated with or without considering sowing density or actual plant density) did not provide the information generally claimed in the literature (i.e. whether a crop is dominant or dominated). Consequently, their use is clearly unadvisable except when analysed jointly with partial land equivalent ratios. The LER index proved to be clearly relevant, versatile and helpful in illustrating the pattern of competitive outcomes in intercropping experiments, in particular when plotting partial LER values of wheat as a function of those of pea. However, LER cannot identify intraspecific and interspecific interactions. To do so we suggest using the intraspecific and interspecific interaction indices, which can also reveal possible facilitation phenomena and allow description of species change in the contribution index (CC). Interaction dynamics between crops that determine the final balance and the outcome of all competitive interactions occurring between the two crops can be evaluated using the CGR index, which is preferable to REIc, particularly when crops differ greatly in their dry weight. Careful choice of index and interpretation of the results are thus essential in correctly understanding species interactions (globally and dynamically) and intercrop efficiency compared with sole crops. Such indices can help highlight and reveal cereal and legume traits suited to intercropping and also appropriate cropping sequences and management techniques, allowing efficient intercropping. However, the results must always be related to actual data values (yield, dry weight or N accumulated) because the indices used cannot evaluate intrinsically quantitative performance but only the relative performance of intercrops compared with that of sole crops.
