D. Mummery et M. Battaglia, Applying PROMOD spatially across Tasmania with sensitivity analysis to screen for prospective Eucalyptus globulus plantation sites, FOREST ECOL, 140(1), 2001, pp. 51-63
When screening areas as prospective plantation sites two criteria are of in
terest: the expected yield at that site and the uncertainty associated with
this yield expectation. This paper describes a methodology to enable spati
al estimates of uncertainty to be attached to predictions of tree growth in
relation to variations in soil depth, nutrient status, and drainage. An en
vironmental sensitivity analysis was applied using the growth model PROMOD
to produce site productivity and suitability maps for Eucalyptus globulus (
Labill.) on a 1000m regular grid of Tasmania. Nine productivity surfaces we
re computed, one for each of three assumed soil-water holding capacities (f
or soil depths of 0.5, 1.0, and 1.5m - the common classes in Tasmania), com
bined with three permutations of soil-nutrient rating (the expected rating
and one class above and one below this value), and incorporating potential
waterlogging estimates. The mean and coefficient of variation was calculate
d for each 1000m grid cell in the nine subsequent productivity surfaces. A
final plantation suitability grid was then mapped to show areas of the Stat
e with predicted high (>25 m(3) ha(-1) per year), moderate (15-25 m(3) ha(-
1) per year), and low productivity (<15 m(3) ha(-1) per year), each with co
rresponding high variability (coefficient of variation (CV>30%)) and low va
riability classes (CV<30%). As no areas had both, high productivity and hig
h variability, only five plantation suitability classes were mapped.
Those areas indicated as high productivity/low variation sites have a high
probability of being successful E. globulus plantations. Low productivity s
ites with either high or low variation are very unlikely to be favourable f
or this species. Moderate productivity sites with low variation are likely
to be viable potential plantation sites with relatively low risk, particula
rly with nutrient amelioration. Moderate productivity sites with high varia
tion might warrant further field investigation to determine if there are ma
jor environmental factors or combinations Of these, such as soil depth (wat
er holding capacity), nutrient status, or drainage that are actually going
to be limiting reasonable growth at these sites. (C) 2001 Elsevier Science
B.V. All rights reserved.