If you are presenting results of soil loss in a numerical value (e.g. in kg of soil lost to the area considered per unit of time (day, month, year)), and have several numerical values for this parameter, then you should be able to undertake a statistical test to assess the significance of the results. Even if you only have one value, publicising your work in a paper is useful for other as long as you detail the experiment, method of interpretation in an objective and scientific manner. If people like the experiment, you may get some funds to do more and be able to "validate" your results i.e. find relevant factors in crop harvesting that affect soil loss!
My Dear Dennis Edem; In general two to three year crop data can give reliable results and also chances to get published in an standard research journal which have good impact factor. Thanks