Ecological Climatic Regions And Tree Distributions In Michigan (sucrose Inversion).

Abstract

Climatic statistics and their relationships to tree distributions were used to identify areas relatively similar in their potential for plant growth. Major objectives of the study were to: (1) compute climatic statistics with potential relevance to plant growth and survival, (2) relate those statistics to the distribution of tree species native in the state, and (3) delineate ecologically similar climatic regions and districts. Statistics relevant to plant growth were computed from daily data, 1951-1980, for 125 weather stations. Major statistics included average annual extreme minimum temperature, extreme minimum temperature for 1951-1980, potential evapotranspiration, ratio of precipitation to potential evapotranspiration, growing season length and heat sum, heat sum prior to last spring freeze, and ratio of night heat sum to total heat sum. Standard deviations were computed and principal component analysis was used to summarize additional climatic statistics. Local regression analyses were used to extend climatic statistics to areas in the state without weather stations. In addition, the sucrose inversion technique was used to obtain an independent estimate of growing season temperatures for the summer of 1983. The occurrence of species in a stratified random sample of forest sites was related to the climatic statistics through histograms, correlation coefficients, Kolmogrov-Smirnov tests, and a binary discriminant analysis ordination. Cluster analyses and discriminant analyses were used to define regions and districts. Species showed individualistic relationships to the climatic statistics. Southern deciduous species tended to be most strongly associated with long growing seasons and high heat sums, high extreme minimum temperature, low mid-summer ratio of precipitation to potential evapotranspiration, and high ratio of night heat sum to total heat sum. Relative frequency of species restricted to the southern part of the state (e.g. Sassafras albidum (Nutt.) Nees) was correlated with low heat sums prior to last spring freeze and low variation in summer warmth variables. Fagus grandifolia Ehrh., showed strong associations with low continentality index and low variation in winter temperature, and little association with measures of growing season warmth. Northern conifers (e.g. Tsuga canadensis L.) appeared tolerant of extreme winter cold, and cool variable growing seasons. Three regions, twenty districts, and one metropolitan subdistrict were delineated.