Michel Genard of INRA Avignon developped a peach tree modle that takes in account several features of quality and climate http://www.jhortscib.com/isafruit/isa_pp157_163.pdf , For tropical fruit crop a specific calibration is required, but mango fruit allocation was modelledin a separate model, as in http://treephys.oxfordjournals.org/content/25/5/583.full.pdf by the same group.with Mathieu Lechaudel of Cirad, but a lot would depend in case of tropical fruits on management of water and fruit thinning.
During the last 20 years, we have been developing a crop model (SIMPROC MODEL) oriented to the evaluation of climatic suitability for several cultivated species. It is an ecophysiologic based model, which simulate weekly dry matter production and cabohydrate partition among organs. Growth rate depend on solar radiation, temperature and evapotranspiration. It it sensitive to frosts, water deficit, temperature stress, cold stress. The model has four versions: annual crops, fruit trees, graslands and forest species. One advantage of this model is that requires few climate data (just to feed internal climate generator) and can iterate during the year in order to look for the best sowing date in the case of annual crops. In the case of fruit trees (SIMFRUT) you have to inform the model the date of bud break (for temperate species) and flowering for tropical species
We have calibrated the model for several temperate species (apples, peaches, pears, plums, table grape, wine grape, almond, bluberries and several others). For tropical species like citrus, mango and guava we just have some calibration for oranges. Maybe you could be interested in calibrate the model for theese species. One of the advantages of the model is that calibration process is not so difficult. The other advantage is that the model can run for a large number of location in orden to eassily map its results. You can see it at: