This paper briefly describes the underlying assumptions and inherent drawbacks of logarithmic regression models commonly used to develop biomass prediction equations. Superiority of simple nonlinear models over both the logarithmic and multiple linear regression models is discussed and demonstrated on two data sets.Ce document decrit brievement les suppositions sous-jacentes et les inconvknients rattaches aux modeles de regression logarithmique employes pour developper des equations de prkdic-. tion touchant la biomasse. L'auteur 6tudie et demontre, au moyen de deux series de donnees, la superiorite des modeles simples non linkaires tant sur les modeles de regression' logarithmiques que lineaires multiples.
This paper presents a growth analysis of a forest drainage experiment carried out over 40 years in northern Ontario. It is based on remeasurement data obtained in 1969 from 38 growth plots established following drainage in 1929 and from increment cores and sectioned trees.Results indicate that both annual tree diameter and height growth increased significantly after draining; that tree growth before draining was related to site quality only, while after draining it was related also to tree vigor and distance of water flow from the nearest ditch; that both stand diameter and height growth were related to site index, stand age, and initial stocking; and that stand basal area and volume growth were, in addition, related to a product sine function of distance of water flow from the ditch, peat moisture, decomposition, and depth.Both individual tree and stand growth responded well to draining, with younger and more vigorous trees that were growing on better-quality sites showing the greatest response. For a given site, growth response was not greatest for trees and stands nearest the ditch, but for those some distance away.
RCsumCA four-parameter sigmoidal growth model was employed to Un modttle de croissance sigmoidale 5. quatre parambtres a express various characteristics of interest in Plonski's metric Ct C utilist pour exprimer diverses caractkristiques intkressantes yield tables as functions of stand age and site index. Except in des tables mCtriques de rendement de Plonski, en fonction de a few cases, excellent fits were obtained. The resulting equa-1'8ge du peuplement et de la qualitt du site. A l'exception de tions will facilitate direct computer use of Plonski's metric yield quelques cas, une adtquation excellente ont Ct C obtenue. Les tables.Cquations qui en rksultent faciliteront l'utilisation directe par l'ordinateur des tables mCtriques de rendement de Plonski.
This paper describes the spatial pattern, expressed by Pielou's nonrandomness index, of trees within 13 sampled tracts from the major forest types of northern Ontario. Results indicate that: (a) the majority of natural coniferous or mixedwood stands have highly clustered patterns; (b) hardwood stands or the hardwood component of the mixedwood stands show nearly 'random' spatial patterns; and (c) uniform spacing in natural stands is very rare. Results also indicate that spatial patterns vary considerably during stand development for the various species group – size class combinations within a stand and between different forest types. The importance of spatial pattern and stand population dynamics is pointed out with regard to growth simulation modelling and mechanized harvesting and thinning studies.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.