7. Conclusions

Willemet) and buffelgrass (Pennisetum ciliare (L.) Link) in Oklahoma, Texas, and Mexico [18]. To evaluate the ability of the model to simulate introduced or improved grasses, we tested coastal bermudagrass (Cynodon dactylon (L.) Pers.) and bahiagrass (Paspalum notatum Flügge var. saurae Parodi) at several sites in Texas [19]. Western grasses in low-rainfall sites in Montana were simulated using parameters derived for some common native grasses there [20]. The cool-season forage "tall fescue" was simulated at several sites in several states where this grass is commonly grown [14]. In addition, creosote bush (Larrea tridentata [DC.] Cov.) parameters were derived the and model testing for its ability to describe competition of this woody species

Overall, the ALMANAC model predicted forage yields with reasonable accuracy, and hence when fully calibrated, the model can be used as an effective management tool to evaluate management practices that maximize forage yields, optimize inputs, and minimize negative

The ALMANAC model uses the best plant growth modeling functions currently developed. Often, knowledge gaps force model developers to use placeholder functions with the hope that future research will enable development of improved, more realistic functions. Some areas for beneficial future research include nutrient and carbohydrate cycling, forage regrowth follow-

The simulated cycling of nutrients in the soil and between the roots to the shoots for perennials needs to be critically investigated for this model. As forages mature and leaves senesce during the fall and winter, often nutrients and carbohydrates are translocated back into the root system, to be used for regrowth the following spring. Grazing may also trigger plants to allocate more carbohydrate storage in roots to survive grazing pressures. Functions describing these processes need to be better developed and incorporated into the ALMANAC model in

Likewise, the regrowth of forages following hay cutting or grazing within the growing season, needs to be more extensively tested. The functions currently in ALMANAC appear to function reasonably. However, as more extensive data are available for testing the model, improve-

The response of forages to applied nutrients often is highly dependent on what is already in the soil. This includes nutrients readily available and those coming from transformations within the soil during the growing season. Very often publications report a nutrient response of a forage without adequately describing initial soil conditions. If adequate nutrients are already present in the soil, the response of the forage to applied nutrients can be much dampened. Likewise, if the soil is initially very nutrient poor, the forage may show a large response to applied nutrients. An extensive testing of the model with data having good values

6. Knowledge gaps and areas for future improvement as a guide for

with forages in arid sites in western Texas [34].

ing haying, nutrient response functions, and legacy effects.

environmental outcomes.

48 Forage Groups

additional research

the future.

ments likely will be made.

for initial soil nutrients will be valuable.

In this chapter, we described the ALMANAC model, including the process simulated, how to derive plant parameters for additional forage species, and how to validate using measured field data. Because of its accurate simulation of plant production, the water balance, and the nutrient balance, the model is capable of simulating a wide variety of environmental and management impacts on forage production, soil health, and conservation concerns, including nutrient and sediment losses. The model will be a useful and valuable tool for forage management in pastures and rangelands in a wide range of conditions.
