Modélisation de la croissance en hauteur dominante et fertilité des peuplements de pin d'Alep (Pinus halepensis Mill.) en Tunisie

EnglishThe Aleppo pine (Pinus halepensis Mill.) is a rus tic species, very present in Mediterranean regions (Parde J. 1957; Quezel P. 1986; Mezali M. 2003). Representing more than 56% of the total forest area (Sghaier 2005), this species is of a great economical, ecological and social values in Tunisia. The development of management models for various area of this species is currently one of the first priorities of the forest administration in the country. It is within this framework that an integrated study on Aleppo pine was undertaken in Tunisia by installing 348 temporary plots with a circular of 400 m2 each. These plots were set in such a way to cover the whole distribution area of the species in the country, from the arid to sub-humid bioclimates. To choose a model which describes the dominant t height growth according to the age and to identify the sites quality relating to this species, varied mathematical models with common asymptote, based on the difference equations method (Table 1), were adjusted and compared. The difference equations method supposes that the observations on a given plot should belong to the same site index curve. The dominant height H2 measured at the age t2 is expressed as a function of t2, the height H1 measured at the age t1, and t1. The expression is obtained by the substitution of one parameter in the growth model height-age (Elfving and Kiviste 1997). The data, which were used for the different adjustments, come from the stem analyses. To estimate the parameters of the various tested models, we performed a non-linear regression using the NLIN procedure of SAS software. A certain number of criteria were used to compare these models (Table 2). These criteria are based on the value and the distribution of the residues and the relation between the observed and estimated values. After analysis, it appeared that the polymorphic difference equation derived from the Lundqvist-Korf function resulted in the best compromise between biological and statistical aspects, producing the most adequate site index curves. This equation is as follows: H2= 39,018 x (H1 / 39,018) (t1/t2)0,351 where H1 is the height (m) at age t1 (years), H2 is the height (m) at age t2 (years). Thanks to this model, the various Aleppo pine stands in Tunisia were divided into 4 classes of quality or productivity. These classes are characterized by average dominant heights obtained at 50 years age of 13,5 m; 10,5 m; 7,5 m and 4,5 m, respectively for the first, second, third and fourth class of quality. francaisPour caracteriser la croissance en hauteur dominante du pin d'Alep (Pinus halepensis Mill.) en Tunisie, differents modeles issus des equations differentielles de Sloboda et McDill-Amateis, ainsi que de la forme integrale de Chapman- Richards, Lundqvist-Korf et Logistique, ont ete ajustes et compares. Les donnees utilisees pour la modelisation sont issues de 348 analyses de tiges d'arbres abattus, soient 1 274 couples hauteur-âge mesures sur des placettes temporaires distribuees sur toute l'aire de repartition de l'espece dans le pays. Des analyses qualitatives basees sur le realisme biologique des modeles, aussi bien que des analyses numeriques et gra phiques basees sur la precision des modeles ont ete employees pour comparer les differents modeles testes. L'equation polymorphique en differences algebriques derivee de la forme integrale de la fonction de Lundqvist-Korf a donne le meilleur compromis entre les aspects biologique et statistique, fournissant les courbes de croissance les plus adequates. Cette equation invariable quant a l'âge de base a permis de repartir la pinede tunisienne en 4 classes de qualite. La premiere classe de qualite est caracterisee par des hauteurs dominantes a l'âge de 50 ans de 12 m et plus. La deuxieme classe pos sede des hauteurs dominantes au meme âge comprises entre 9 et 12 m. La troisieme classe englobe les peuplements dont les hauteurs dominantes a l'âge de reference varient entre 6 et 9 m, et finalement la quatrieme et derniere classe de qualite dont les hauteurs dominantes ne depassent pas les 6 m. Environ la moitie des peuplements echantillonnes appartient a la troisieme classe.