The limitations of and potential solutions to modeling nut tree growth

Nut trees have strong economic impacts. However, these and other crop trees may use much more water than other row crops. Additionally, the current efficiency of fertilizer use for nut trees is only 50%, and fertilizer prices are increasing. Consequently, efficient irrigation and fertilization strategies are needed, and nut tree growth modeling is an effective tool for developing these strategies. However, there are some limitations of current nut tree growth models. Nut tree growth models fall into two basic categories: mechanistic and functional. Mechanistic models are more valuable to the developer for synthesizing research understanding; however, mechanistic models are more difficult to develop. Functional models are robust and easy to understand and run, but are not necessarily accurate outside the environmental conditions that were used to develop the models. The purposes of the model (research or application) determine which kind of model should be used. A model can be built around object-oriented programming, where objects may be added or subtracted from the model with ease. In the past, tree growth models were written with subroutines to represent objects, but the structure was such that they could not be easily removed because the attributes of the object were not part of the subroutine but were a separate object. Another problem is that traditional models are usually written in FORTRAN or some other computer language not understood by the general scientific community. With increases in computing speed, spreadsheet programs can replace the FORTRAN-based crop simulation models as long as the models are simple in structure. The basic principle of tree growth models to control maximum yield is controversial. Some modeling researchers use carbohydrate reserves in the tree root to control the flowering and then the yield, and this approach is indirectly supported by experimental data. However, there is no direct data to prove this principle. Therefore, direct experiments about the relationship of carbohydrate reserves, flowering, and yield are needed. Much research has examined tree growth under ideal conditions when there is no water or nitrogen stress. However, tree growth models need valid equations for when the trees are under stress. A literature search yields only controlled nitrogen-stressed experiments for nut tree seedlings. Thus, nitrogen- and water-stressed experiments for mature trees are needed. In addition, the two stressors should be tested to determine if they are addictive or multipliable. Accurate tree root depth data are also needed for tree modeling because current models only assume certain values for root depth.