The soil biologist has two important roles in the application of soil biological knowledge to sustainable land management. First, soil biologists have a responsibility to continue improving knowledge about biological processes and organisms in soil. Second, they have to apply a rigorous experimental approach when assessing how land management practices affect soil biological fertility. The significance of experimental results is interpreted in terms of the statistical probability of certain occurrences. Although soil biological information is only one component of the knowledge-base required for identifying sustainable land management practices, the lack of attention to this area in recent years means that special focus is required now to redress this deficiency.
When an experiment is designed a number of points need to be considered so that experiments produce valid and accurate results and so that the results are transferable, durable and useful. Scientific studies that are deficient in their design are wasteful of resources and fail to contribute to the advancement of knowledge.
Scientific experiments test hypotheses that are developed according to either prior knowledge or ‘best guesses’. Hypotheses can be supported by the results of an experiment, but not proven. Data from different studies can be collated (via meta-analysis) to gain greater support for an hypothesis.
Experimental design has treatments that are compared to determine whether the results support a particular hypothesis. In addition, sampling and other methods as well as replication all need to be selected precisely so that the experimental effort expended is not wasted.
Sampling: The number of samples taken and the method of taking them are both important. For example, consistency in depth of soil sampling is important for comparability between sampling points. If the question being investigated concerns the presence of organisms in the surface layers of the soil compared with deeper in the profile, the depth of sampling needs to be appropriately selected.
Treatments: In scientific experiments, treatments are the factors that vary. Examples of treatments are types of organic waste and soil temperatures. Complex experimental designs with numerous treatments varying simultaneously are difficult to interpret. Experiments that are relatively simple and address one hypothesis at a time are most appropriate but may be insufficient to explain complex soil biological relationships. Although this is a limitation of the scientific method, over time a substantial amount of information is accumulated that can be interpreted by a land manager.
Duration of study: The time over which an experiment is run needs to be selected so that the hypothesis is adequately tested. Often, short-term experiments provide little information because organisms take a long time to establish stable communities capable of influencing processes such as nutrient cycling soil or soil aggregation. The duration of field experiments is especially important for studies where different land management practices are compared. Results from short-term experiments may not be interpreted in the same way as long-term effects of the same land management practices.
Comparability between sites: Studies at one site are usually not sufficient for identifying the general impacts of land management practices. Experiments conducted at several sites are most useful but the placement of treatments within an experiment needs to be considered carefully due to possible heterogeneity in soil physical and chemical characteristics across a site.
Replication: Within each experiment, replication needs to be sufficient to allow a thorough statistical analysis to be carried out. A minimum of three, four or more replicates of each treatment is essential for all experiments. More replication is required when the plants or soils used are very heterogeneous, as often occurs in natural ecosystems.
Databasing and archiving: Data sets can be combined and analysed with powerful models.
Scientific investigation can support or refute claims for the beneficial impacts of certain practices. This includes the application of commercial products containing living organisms. However, there are many questions that need to be asked that are difficult to answer using experiments in the area of soil biology. This is due to the difficulty in assessing some biological processes, the difficulty in identifying organisms in soil, and the microscopic scale at which much of the activity takes place.
Finally, there are additional areas of investigation that impinge on soil biological fertility that have not yet been addressed scientifically. Land managers need to collaborate with soil biologists as well as with other soil scientists to identify which research areas will lead to a greater understanding of the complex effects of land use on soil biological fertility.
• The scientist has the responsibility to gain knowledge of biological processes and organisms in soil.
• The scientist applies an experimental approach to assess how land management practices are likely to enhance or reduce soil quality.
• The scientist needs to understand how to clearly define and justify (i) soil sampling procedures, (ii) selection of treatments, (iii) duration of the study and (iv) suitability of comparisons between sites.
• The scientist needs to be aware of the high degree of variability within a soil type in the field, and between different soil types because not all experimental data are transferable between sites.
• The scientist provides a framework to be used by land managers who will integrate knowledge from many sources (scientific, social, financial) when making their decisions.