Julia Touza-Montero and Charles Perrings

1
Policies for the management of landscape
diversity and collectively managed forests the
case of Galicia
Julia Touza-Montero and Charles
Perrings Environment Department, University of
York
2
Presentation overview

• Motivation
• Multiple stand landscape model - optimal
  harvesting rule -
• Case study Galician collective forests in Spain
• Policy implications

3
Motivation

• Management of forest resources has moved towards
  a
• landscape-based approach to manage for multiple
  values
• Think beyond individual stands -- mosaic of
  stands
• Why?
• Management at a stand level impedes the
  assessment of the implications of the management
  actions at a landscape scale
• Lack of spatial consideractions (stands size,
  shape, proximity, dispersion, adjacency)
  Unanticipated ecological changes

4
Motivation - Cont.

• Forest management at a landscape scale is
  increasingly recognised as a key for conservation
  of biodiversity
• Focus on maintaining the habitats
• Interactions between spatially dispersed stands
  determine forest ecological processes (e.g.
  movement of species, spread of disturbances)

5
Motivation - Cont.

• Case study focuses on Galicia (Northwest of
  Spain)
• Forest area covers 69 of the territory
• Pinus pinaster, Eucalyptus globulus, and mixed
  forest of
• these species 35
• Quercus robur and Quercus pyrenaica 14
• Non-wooded (scrublands) 31
• Galicia has patchy forest land

6
Motivation - Cont.

• Individual ownership 68 of forest land- 2.3 ha
  mean size
• Collective ownership 30 of forest land- 231 ha
  mean size
• Members of a rural community have the rights to
  the forest resources without parcelling the
  rights to the forest itself
• Forestry, support for cattle raising and
  agriculture, amenities, hunting, etc.
• Multiple forest uses and spatial interactions
  between stands are integrated within the decision
  making process

7
Multiple stand model

• Harvest decisions, i.e. rotation periods, in a
  multiple stand forest managed for timber and
  non-timber values
• Stand interactions are assumed to influence the
  flow of non-timber benefits provided for the
  entire forest landscape
• Bowes and Krutilla, 1985, Swallow et al. 1997,
  Thavonen and Salo, 1999, Amacher et al. 2002
• A dynamic optimal cutting rule in a multiple
  stand forest landscape
• Note theoretical results independent of type of
  ownership

8
Multiple stand model
9
Multiple stand model

• Choose the optimum time moments of the stands
  harvests, to maximise timber and non-timber
  benefits

T
? ? (s1a1(t),..,snan(t)) e-?t dt ??
pixi(?ij-)-cipxip e- ??ij subject to xi
Fixi(t) between harvests ai 1
between harvests xi(?ij) - xi(?ij-)
- xi(?ij-) xip at harvest ai(?ij) -
ai(?ij-) - a(?ij-) aip at
harvest xi(0)x0
0
?
?
10
Multiple stand model

• Optimal cutting condition for any stand in the
  forest landscape
• Forest MB Forest MC

?(s1a1(?ij-),., snan(?ij-)) e- ??ij -
?(s1a1(?ij),.., snan(?ij)) e- ??ij pi
Fixi(?ij-) e- ??ij
? pixi(?ij-)-cipxip e- ??ij pi
Fixi(?ij1-) e- ??ij1 ?i(?ij)
?(?ij-) - ?(?ij) stand i relative contribution
to the forest NTB if its harvest is delayed
Stands interactions
?i(?ij) opportunity costs of delaying future
forest NTB and altering the age of
stand i relative to the other stands age
11
Case study Galician collective lands

• Analysis of harvesting decisions accounting for
  landscape pattern differences between collective
  forests
• Dependent variable rotation length
• Baixo-Miño (68 forest land 73 collective
  forests)
• Reports on clear-cuttings undertaken on
  collective forests under contract mechanisms -
  from 1995 to 2001

12
Case study Galician collective lands
Duration analysis parameter estimates from the
Weibull distribution
Coefficient p-value Coefficient p-value
N 108a N 94
DV Conifer 0.543 0.000 0.580 0.000
Net timber prices 0.582 0.003 0.607 0.060
Annual difference in interest rate (IRt-IRt-1) -0.132 0.006 -0.119 0.032
Mean size of wooded patches 0.465 0.005 0.427 0.018
of area occupied by mixed tree-species patches 0.436 0.000 0.401 0.001
Clusters of patches of identical forest characteristics -0.192 0.033 -0.206 0.044
Constant 2.528 0.020 2.215 0.106
P 2.788 0.000 2.663 0.000
Log-Likelihood -61.80 -57.42
Wald-test (p-value) 0.764 0.766
p-value 0.000 0.000
(a) We compute this estimation with dummies for
those harvesting reports which contain several
tree species but the p-values suggested omitting
them from the model significant at 1 level
significant at 5 significant at 10
13
Conclusions and policy implications

• Optimal harvesting strategies at a single-stand
  scale are not necessarily optimal when a larger
  spatial scale is adoped and spatial interactions
  inform the decisions
• What is the difference?
• Single-stand the flow of the stands NTB
  influences when the
• stand should be harvested (Faustmann-Hartman
  rule)
• Multiple-stand it is the relative contribution
  of each stand to the NTB from the overall forest
  landscape that affects the rotation intervals

14
Policy implications Cont...

• Fragmentation, diversity and clumpiness are
  relevant determinants of harvesting behaviour in
  Galicia collective forests
• Policy implications
• Non-harvesting policies may be optimal for
  those areas that
• contribute highly valuable environmental goods
  and services to
• forest landscape benefits
• It may be optimal never to harvest any of the
  stands in the forest
• if NTB increase with age and are significant with
  respect to other
• uses

15
Policy implications Cont...
The ecological and economic consequences of
alternative actions taken at small scales
(stands) on a wider spatial context (i.e.
forest landscape) must be allowed for in forest
management decision making The weight attached
to the forest benefits from a particular
stand depend on the interdependence between
stands Importance of spatial interactions on
harvesting strategies Scale
16
Policy implications Cont...
98 of forest land is in private hands Individual
ownerships have a mean size of 2.3 hectares
In Galicia Public planning of forest
landscapes is essential to account both for (a)
spatial interactions between the stands (b)
preferences of society Public policies, which
encourage coordination and cooperation among
forest owners, are necessary to ensure that
private owners actions are consistent with
environmental and economic goals set at a
landscape level Problems Inventories of single
and collective forest ownerships are scarce
Weakened communal institutions
17
Thank you!