Discussion

Using plants as a filtration system for aquaculture effluents has been well documented, with approaches ranging from the use of constructed wetlands (Greenberg 1991; Redding et al 1997; Rosati and Respicio 1999) to the incorporation of aquaculture effluents into hydroponic vegetable production (McMurtry et al. 1990), with the majority of the botanical approaches to aquaculture effluent treatment being aimed at freshwater systems. While the value of mariculture products is greater than that of freshwater aquaculture products (FAO 2000), botanical approaches to the treatment of mariculture effluents are not as well developed as those for freshwater aquaculture effluents, being limited to seaweeds (De La Cruz 1994; Troell et al. 1999) and halophytes (Brown and Glenn 1999; Brown et al. 1999).

McMurtry and co-workers (1990), among others, have reported that aquaculture effluents from freshwater production system can supply all of the necessary mineral nutrition for vegetables, including bush beans (Phaseolus vulgaris), cucumbers (Cucumis sativus) and tomatoes (Lycopersicon esculentum). It seems logical, then, to assume that with the successful production of various marine species in low-salinity waters (Cawthorne et al. 1983; Fosberg et al. 1996; Fosberg and Neill 1997; Flaherty and Vandergeest 1998; Samocha et al. 1998) and the salt tolerance of certain crops, that integrating low-salinity mariculture and agriculture could result in similar findings.

In the current study, statistically significant differences were found in respect to total nitrogen levels in the irrigation water, with the 100% well water (negative control) treatment having the highest nitrogen levels. However, no statistically significant differences were found in tree growth as measured by either height or diameter among the three treatments. Looking at Figures 4 and 5, we can see that slope of the lines fitted to the data corresponding to each treatment are not the same, suggesting that while growth of trees was not statistically significant over the four month study period, a longer term study might show significant differences. In Figure 4, the slope of the line fitted to the 100% well water treatment data is less than the line for either the normal management treatment or the 100% effluent treatment. Figure 5 suggests that, when extrapolated out, growth of the normal management treatment would be greatest, followed by the 100% well water treatment followed by the 100% effluent treatment.

Our findings are in clear contrast to what has been shown in other research for vegetables (Prinsloo and Schoonbee 1987; McMurtry et al. 1990), seaweeds (Troell et al. 1999) and halophytes (Brown and Glenn 1999; Brown et al. 1999). Reasons for the apparent discrepancies could be the increases observed in soil salinity or, more likely, the high levels of nitrate present in the well water. The farm in the study is located in an area that has a long history of being planted in cotton. Our belief is that potential benefits of irrigating with nutrient rich, low-salinity shrimp farm effluents have been masked by the high nitrate-nitrogen levels present in the well water. While irrigating with 100% low-salinity effluent did not improve growth of olive trees in respect to either the normal farm management or the 100% well water (negative control) during the four month study, our results do indicate that irrigating with this low-salinity shrimp farm effluent had no noticeable negative effects.

Despite the promising results obtained in this preliminary study, many problems associated with conducting research at a commercial farm were encountered. Over the four month study period, the planned irrigation schedule was interrupted on five separate occasions due to either miscommunication and/or logistical difficulties. While these relatively minor setbacks are to be expected on a commercial farm, they do call into question the scientific integrity of the data obtained. It is for this reason that we are planning to continue this research for another season.