Record Details
A preliminary vegetation classification of the Tombstone, Arizona, vicinity
ScholarsArchive at Oregon State University
| Field | Value |
|---|---|
| Title | A preliminary vegetation classification of the Tombstone, Arizona, vicinity |
| Names |
Garcia-Moya, Edmundo, 1941-
(creator) Poulton, C. E. (advisor) |
| Date Issued | 1972-03-06 (iso8601) |
| Note | Graduation date: 1972 |
| Abstract | The need for classifying vegetation in a more precise way is evident. Also, there is a need to provide a hierarchical classification scheme that will match changes in image characteristics as one moves through the scales from space to conventional aerial photography. Such refined classifications of vegetation are the first steps toward a better understanding of the potentialities and limitations of a specified area which help in the detection of analogous environmental conditions for resource allocation and management purposes. These needs become more urgent as use and competition for natural resources and land increases. The first approximation of a classification scheme may meet these needs for a test site in the Tombstone, Arizona, vicinity. This classification task was accomplished by the use of a "hierarchical- polythetic-agglomerative" package using presence-absence data and standardized cover estimates, The following tentative associations and a variant were found upon division of the original data into groups of convenient size to meet the limitations of the programs: Association A (Panicum hirticaule/Tidestromia lanuginosa- Boerhaavia coulteri) la (a variant of Association A) Association B (Rhus microphylla-Dalea formosa) Association C (Gutierrezia sarothrae/Eriogonum abertianum) Association D (Menodora scabra/Tridens grandiflorus) Association E (Hilaria belangeri) Association F (Gilia rigidula-Rhynchosia texana) Association G (Hilaria mutica/Eriochloa gracilis/Crotalaria pumila) Association H (Haplopappus tenuisectus /Eragrostis lehmanniana) As sociation I (Ayenia pusilla/Eragrostis intermedia) Association J (Cnidoscolus angustidens) Association K (typical association lacking character species) of Alliance III (Fouquieria splendens-Acacia constricta-Aloysia wrightii) Association L (Agave palmeri-Agave parryi/Haplopappus laricifolius) Association M (Mortonia scabrella) These were grouped by the classification into units of higher rank. Association (A), (B), and the variant (la) grouped into an alliance with the character species, Acacia vernicosa-Larrea tridentata-Flourensia cernua, Associations (C), (D), (E), and (F) grouped into the Yucca elata/Bouteloua eriopoda Alliance; and Associations (I), (J), and (K) were held together in an Alliance by the character species, Fouquieria splendens-Acacia constricta-Aloysia wrightii. Validity of the character and differential species as association and subassociation "identifiers" was tested by the use of stepwise discriminant analysis. Two hierarchical levels and two runs per group were carried out using this analysis. A perfect discrimination between or among the groups was found at all levels except one. Species identified as differential or character species in the classification process were found to be the best discriminants. This suggests that the vegetation units identified by these species are valid. The "hierarchical-polythetic-agglomerative" approach to vegetation classification provides the kind of framework which is compatible with the classical phytosociological techniques of vegetation study. It eliminates, however, the cumbersome task of hand sorting and tabulation and increases the capacity of complex operations as well as introduction of more systematic and thorough evaluation into the analysis. This approach of classifying vegetation appears to be suited for survey-type studies in areas where vegetation information is limited and the need exists for an initial classification in order to begin more comprehensive quantitative studies. This does not preclude using the method for other purposes. Because it is a hierarchical method, one can go into as much classification detail as is dictated by the purpose of the vegetation study. This last feature is well suited to the use of the results in interpretation of multiscale photography so important in the field of resource analysis. Research needs to be done to answer the very fundamental question of why Euclidean distance (as a measure of similarity) and Ward's method (as a sorting strategy) provided a more adequate hierarchical classification scheme when presence and absence data are used rather than standardized values. Research is needed on the most effective criteria to divide large data sets into groups of the appropriate size. Results of this classification now need to be tested by practical field use in the recognition of ecosystems and their mapping on appropriate scales of remote sensing imagery. To aid the practical user of this information, a dichotomous key to the phytosociological classes was developed. This requires the recognition of 8 species to make classifications at alliance level and 32 species to achieve the association separations. A key for subassociation level was not prepared because the practical value of this level is somewhat in doubt. |
| Genre | Thesis/Dissertation |
| Topic | Botany -- Arizona -- Tombstone |
| Identifier | http://hdl.handle.net/1957/46146 |
