Harvest management - Ecological Applications: Toward a Sustainable World - page 202

Table 7.1 Values used
in a matrix model of a
mixed beech forest in
the South Island of New
Zealand. Trees are
classifi ed into 11 size
classes according to
diameter at breast
height (DBH, a standard
forestry measure).
Growth is expressed as
increases (mm yr −1 ) in
DBH. Recruitment and
mortality rates are per
tree per year. Smaller
trees have zero or small
recruitment rates and
higher mortality rates
than their larger
counterparts. (From
Efford, 1999.)

Trees Grow th

Size class (mm) (per hectare) (mm yr − 1 ) Recruitment rate Mortality rate

0-100 742 2.56 0 0.0748

100 -200 61 2.55 0.029 0.0236

200-300 29 2.85 0.079 0.0075

300-400 25 3.11 0.155 0.0075

400-500 21 3.19 0.257 0.0075

500-600 22 2.88 0.383 0.0075

600-700 15 2.62 0.536 0.0075

700-800 11 2.32 0.713 0.0075

800-900 7 2.04 0.916 0.0075

900-1000 4 1.75 1.144 0.0075

1000 + 6 1.47 1.398 0.0075

Net- and line-fi shing methods, on the other hand, catch fi sh before their size is
known and usually damage them so that subsequent release is not feasible. Marine
protected areas, as long as they have the appropriate properties to contribute to the
harvest (Section 7.2.4), may have the additional benefi t of protecting the really big
mothers.
By way of summarizing the management measures available for fi sheries, you can
envisage these as either input or output controls . Input controls include limiting the
number of fi shing vessels permitted or restricting the size and fi shing power of their
gear. Output controls , on the other hand, include restrictions on the size composition
of the catch, the total allowable catch of fi sh to be taken by the fi shing fl eet, or of
individual quotas for each vessel. Economic checks on catch size, such as govern-
ment taxes on fi sh landed, are also sometimes used as output controls.

7. 3 . 2 Forestry -
axeman, spare which
tree?

Models used to manage harvests in native forests follow a similar procedure to the
dynamic pool fi sheries models, but in many respects the problem is more tractable
- the trees can be counted and measured and do not move around, and often only
a single company is responsible for harvesting them. In the case of a proposal to
sustainably harvest a mixed native beech forest ( Nothofagus spp. and Dacrydium
cupressinum ) in New Zealand, a matrix model was used of the type described in
Section 5.4, in which tree size classes (measured according to diameter at breast
height, DBH) were used instead of age classes. The number of trees in each size class
(0 -100 mm, 100 -200 mm, etc.) changes from year to year as some die or are felled,
and others grow across the boundary to the next size class. Saplings are considered
to be recruited to the smallest tree size class once they attain a height of 1.4 m.
Known rates of growth, reproduction (number of sapling recruits per individual per
year) and mortality were fed into the model (Table 7.1). Finally, a sustainable yield
was estimated by fi nding the mortality rate in each size class that will maintain the
current size composition, and treating that mortality as a harvestable surplus. To
allow for uncertainty, the proposal was to only take 50% of this surplus, at least in
the fi rst phase of an adaptive management approach.
As usual with any model of a complex system, one can argue about the most
appropriate model structure and parameter values. In particular, as Efford (1999)
pointed out, we do not know for certain that harvest mortality substitutes for natural