Comparison of treatments and testing new
Most formal farm forestry trials involve the
testing of hypotheses or the comparison of alternative treatments.
A carefully prepared experimental design reduces natural variability
and allows statistical analysis to test for any significant
difference between treatments. To be credible, a number of
replications of each treatment need to be incorporated to
account for natural variability. The replicates, or plots,
may vary in size from just one tree to many hectares depending
on the treatments being tested. For example, single tree plots
might be used to test for genetic variability within a species
whereas small area plots are commonly used in establishment
trials testing fertiliser, weed control and soil preparation
options. Large plots are required to test silvicultural options
and agricultural productivity.
Replicated trials can be very complicated to
design, manage, measure and interpret. However, farmers can
use scientifically valid methods to undertake simple and inexpensive
tests. The key is to minimise the number of treatments and
the variability across the trial site so that any observations
can be legitimately related to the treatment. For example,
to determine if there is any advantage in ripping before planting
a farmer could establish a replicated trial and monitor tree
growth over the following years. By plotting the results they
can judge whether the treatment is worthwhile or even undertake
a statistical analysis to test for significant differences
between the treatments. The treatment plots should be carefully
located and protected from anything that may have an uneven
effect across the site. To avoid any outside influence or
edge effect a buffer, treated in the same way as the plot,
might also be required.
One interesting trial design commonly used in
agroforestry and farm forestry research is the variable spacing
trial. Conventional spacing trials require a very large area
of land because each plot containing a particular tree spacing
option needs a large buffer to minimise the edge effect. Variable
spacing trials simply involve a gradation from a high tree
stocking down to a low stocking across a uniform site. See
Complex equations are available to help determine
the effective stocking rate of each row of trees. The designs
can be set out as parallel lines or like spokes of a wheel
(Nelder). To be statistically valid there must be at least
7 trees at each stocking and the entire trial should be replicated
at least three times. It is critical that there is a well-grown
tree at every point of the design. Although not strictly valid,
a single trial is useful to demonstrate the effect of spacing
on tree growth and pasture production.
Testing and adapting new ideas
Farm Foresters have a reputation for trying
out new ideas or working to improve old ones. Testing new
or alternative species, management options, equipment, pruning
techniques, and potential forest products in a small way,
without risking large amount of money or time, provides valuable
information prior to a large investment being made. Farmers
commonly share their experiences with others, thereby helping
them deal with similar problems/issues.
Many farmers have also taken to adapting existing
agricultural equipment or building their own new tools for
the establishment, management or harvesting of their forests.
Direct seeders, planting machines, pruning tools and other
farm inventions are commonly seen at field days. Some of the
better ones go on to be commercially produced.
Being part of a regional farm forestry network
or grower group is the best way to gather ideas from others.
Local newsletters, meetings, field days and seminars provide
growers with the opportunity to share ideas and discuss strategies
for regional initiatives. Magazines such as Australian Forest
Growers and Agroforestry News include many articles about
farmer experiences and innovations.
Back to top