Diapositive 1

1
Modeling the distribution of Pinus strobus L. at
the beginning of the 19th century in central
Quebec (Canada) using presettlement land survey
records
Eduard Mauri Ortuno1, Frédérik Doyon2, Alison
Munson1 1 Centre détude de la forêt, Faculté de
foresterie, géographie et géomatique, Université
Laval 2 Université du Québec en Outaouais,
Institut québécois daménagement de la forêt
feuillue
Introduction Eastern white pine (WP) is one of
the most valuable conifers in eastern North
America and part of our ecological heritage. In
the province of Quebec, according to historical
records, by 1900 the most accessible mature
specimens had been selectively logged,
particularly along the St. Lawrence River and its
major tributa-ries, leaving relic populations
that are poorly representative of its
presettlement distribution 1, 2, 4, 5, 8. For
example, in the study area, mature WP prevalence
decrea-sed from 28 in the 19th century (land
survey observations) to a current 14 , while
conserving a high site fidelity 6. In La Mauricie
national park (province of Quebec), WP volume
passed from 12.5 in the 19th century to a
current 0.5 7, 10. Because WP presettlement
abundance is unknown and its present distribution
has been largely reduced, we aim to develop a
tool to guide its ecological restoration.
Methods 19th century WP (WP19) WP
presence/absence is obtained from 13 land survey
logbooks dating from 1795 to 1895. 2614
observation points were extracted.
Study area It is limited by the intersection of
the Mauricie region (province of Quebec) and the
Acer saccharum-Betula alleghaniensis bioclimatic
domain (about 7,000 km2), part of the Great
Lakes-St. Lawrence forest. Landscape is mainly
covered by glacial and fluvio-glacial deposits,
with a relief composed of hills and high-hills,
with moderate and steep slopes. Climate is
continental subpolar, subhumid, with an average
temperature of 3C, 900 to 1,400 mm rainfall, and
160 to 180 growing season days 9.

• Current WP traces (traces20)
• Obtained from 421 randomly chosen land survey
  points
• 222 with WP19
• 199 without WP19
• The plot radius of 20 m was established as the
  depth the surveyor could distinguish the
  vegetation.
• Traces20 were the abundance of
• WP stumps
• Mature WP
• WP saplings
• WP seedlings

Environmental conditions (env) Obtained from
photo interpretation or digital elevation model.
Associations with WP19 The order of abundance of
each taxon was transformed to frequency using the
broken stick model 3. Retaining taxa present in
more than 1 of observation points, a
hierarchical cluster analysis was performed on
the 737 points containing WP19 to create six
cover associations.
env type cat.
Soil deposit Nominal 6
Drainage Ordinal 7
Slope Cont. -
Aspect Nominal 4
Slope position Nominal 3

• Management history
• 1830 1850 beginning of WP selective cuts
• 1850 1890 WP is still the most exploited
  species
• 1890 mature WP depletion 2, 3, 5.

Old white pine stump
Eastern white pine (Pinus strobus L.) The species
is a long-lasting pioneer favoured by surface
fires, which burn the organic horizon and expose
the mineral soil needed for seed germination.
Because of its sensitivity to light competition
at a young age, it is usually relegated to
well-drained, coarse-textured sites 11, where the
cover takes more time to close. Eastern white
pine is present through all successional stages,
among conifers as well as shade tolerant and
intolerant hardwoods, but it rarely forms pure
stands 12.

WP19 f (env)
WP19 f (traces20)
Associations WP19 f (env)
study area survey point with WP19 survey point
without WP19
This represents a state where WP was still within
natural ranges.
20 km
Results WP19 f (env) Drainage
was correlated with soil deposits, and slope did
not add any extra information.

• Discussion
• The model WP19 f (env) is similar to known WP
  autoecology. Its weak prediction is due to the
  high frequency of glacial deposits (66 of
  survey points), which are not statistically
  significant to predict WP19 presence/absence.
  Because glaciolacustrine and marine deposits
  (well drained and coarse textured) are situated
  near the Saint Lawrence Lowlands where settlement
  started, WP would have already been harvested
  there before land surveys were carried out.
• A more accurate model is WP19 f (traces20), but
  it has two main handicaps
• no possible remote sensing for traces20,
• stump decomposition will erase this past
  evidence.
• Using abundance of WP19 and traces20, instead of
  pre-sence/absence did not improve the models
  prediction.
• The frequency of each cover association with WP19
  can be used to guide a more comprehensive
  restoration.
• In the region, major users would be
• Projet Triade forest ecosystem management,
• La Mauricie national park WP restoration
  program.
• Models presented here can be extended to the
  entire ecological region but not to the Saint
  Lawrence valley, due to major differences in
  topography and soil deposits. However, similar
  studies could be carried out.

Associations WP19 f (env) The multinomial
logistic regression was not significant, nor
univariate regression with only one environmental
predictor at a time. WP ecological restoration
tool Probabilities from the model WP19 f (env)
are calculated for a 19 x 19 m cell grid,
covering the study area where environmental
predictors are available. There are 72 possible
environmental combina-tions. Of these
combinations, 60, covering 99.4 of the study
area, are taken into account by the model. 49
of this area, according to the 0.28 threshold,
has a higher chance of WP19 presence than
absence. Probability of WP19 presence ranges from
0.08 to 0.50.
Nagelkerke R2 0.05 Nagelkerke R2 0.05 Threshold 0.28 Threshold 0.28
Predicted Predicted
no yes Total Total
Observed no 1058 819 1877 1877
Observed yes 304 433 737 737
total 1362 1252 2614 2614
Sensitivity 59 overall correct overall correct overall correct 57
Specificity 56 overall correct overall correct overall correct 57
env WP19 prevalence WP19 prevalence estimates
Deposit Deposit
Outwash-Alluvion Outwash-Alluvion 45 0.86
Thin glacial-Rocky Thin glacial-Rocky 36 0.37
Glaciolacustrine-Marine Glaciolacustrine-Marine 12 -1.00
Glacial Glacial 26 Not sig.
Ice-marginal Ice-marginal 27 Not sig.
Organic Organic 20 Not sig.
Aspect Aspect
S-SW-W S-SW-W 34 0.19
NE-E-SE NE-E-SE 24 -0.25
NW-N NW-N 27 Not sig.
Flat Flat 28 Not sig.
Slope position Slope position
Top-High slope Top-High slope 35 0.29
Mid-Bottom slope Mid-Bottom slope 27 Not sig.
Valley-Flat Valley-Flat 23 -0.22
Overall Overall 28
Steven Katovich, USDA Forest Service

• Objective and hypothesis
• Understand the presettlement distribution of WP
  in the central Mauricie region (province of
  Quebec), by relating past WP presence, as
  described by primeval land survey records, to
  environmental variables and current traces.
• Create a tool to guide the ecological restoration
  of WP, and its forests types, for the central
  Mauricie region.
• The hypotheses are

WP19 f (traces20)
Stumps, seedlings and saplings are positively
related to WP19 presence.
traces20 estimates
Stumps 1.26
Seedlings 1.59
Saplings 1.47
Mature WP Not sig.
Which environmental conditions did WP occupy in presettlement forests? WP19 f (env)
Are current WP traces reliable to describe past WP distribution? WP19 f (traces20)
Which species accompanied WP in presettlement conditions? Associations WP19 f (env)
Nagelkerke R2 0.24 Nagelkerke R2 0.24 Threshold 0.53 Threshold 0.53
Predicted Predicted
no yes Total Total
Observed no 150 49 199 199
Observed yes 75 147 222 222
total 225 196 421 421
Sensitivity 66 overall correct overall correct overall correct 71
Specificity 75 overall correct overall correct overall correct 71
References (1) Bouchard, P. 1999. Étude sur
l'importance relative du pin blanc à l'intérieur
des forêts naturelles du parc national de La
Mauricie. Del Degan, Massé et Associés inc.,
Québec, QC. (2) Brisson, J., and Bouchard, A.
2003. In the past two centuries, human activities
have caused major changes in the tree species
composition of
Acknowledgments
southern Quebec, Canada. Ecoscience 10(2)
236-246. (3) Frontier, S. 1976. Decrease of
eigenvalues in princiapal component analysis -
comparison with broken stick model. Journal of
Experimental Marine Biology and Ecology 25(1)
67-75. (4) Gélinas, C. 1984. L'exploitation et la
conservation forestière au parc national de La
Mauricie 1830-1940 dossier documentaire. Parcs
Canada, Québec, QC.
(5) Lafleur, N. 1970. La drave en Mauricie, des
origines à nos jours histoire et traditions.
1st ed. Éditions du Bien public, Trois-Rivières,
QC. (6) Mauri Ortuno, E., Doyon, F., and Munson,
A. 2008. Poster Modélisation de la distribution
du pin blanc au début du 19e siècle dans la
Moyenne-Mauricie. Colloque Triade. Shawinigan,
QC, september 25-26, 2008. (7) Pelletier, H.
1998. Plan de conservation des écosystèmes
terrestres, parc national de La Mauricie. Service
de la conservation des ressources
naturelles, Parcs Canada, Québec, QC. (8)
Quenneville, R., and Thériault, R. 2001. La
restauration des écosystèmes de pin blanc (Pinus
strobus) - Un enjeu majeur pour le parc national
de La Mauricie. Le Naturaliste canadien 125(2)
39-42. (9) Robitaille, A., and Saucier, J.-P.
1998. Paysages régionaux du Québec méridional.
1st ed. Les Publications du Québec, Sainte-Foy,
QC. (10) Service de conservation des ressources
naturelles. 1998. Cadre la restauration
écologique du pin blanc au parc national de la
Mauricie.
Parcs Canada, Québec, QC. (11) Vlasiu, P.D.,
Nolet, P., and Doyon, F. 2001. Le pin blanc -
Revue de littérature. Institut québécois
daménagement de la forêt feuillue, QC. (12)
Wendel, G.W., and Smith, H.C. 1990. Eastern White
Pine. In Conifers. Edited by R. M. Burns, and B.
H. Honkala. United States Department of
Agriculture, Forest Service, Washington, DC.