ISSN:
1573-5036
Schlagwort(e):
aerial photography
;
GIS
;
image analysis
;
microvariability
;
millet
;
Sahel
;
spatial variability
;
topography
Quelle:
Springer Online Journal Archives 1860-2000
Thema:
Land- und Forstwirtschaft, Gartenbau, Fischereiwirtschaft, Hauswirtschaft
Notizen:
Abstract Variability in plant growth is high on most sandy soils of the West African Sahel, often requiring extensive destructive sampling for the reliable estimation of treatment effects. A non-destructive method using aerial photographs and topographic measurements integrated in a Geographic Information System (GIS) was evaluated to determine the effects of organic and inorganic soil amendments on the growth of millet [Pennisetum glaucum (L.) R. Br.] and Guiera senegalensis J.F. Grmel. Based on aerial photographs, quantitative methods were developed to estimate the dry matter of millet plants and Guiera coppices present in the field each year prior to millet sowing. Integrating digital images of both plant species, measurements of the field's topography and a map of the experimental layout in a GIS allowed successful monitoring of the growth of both species as influenced by phosphorus application and the shrub-crop interaction. Regressions between the dry matter of Guiera coppices and the canopy area were good (r = 0.76 to 0.93) and permitted the calculation of the individual coppice dry matter for the entire field with fewer than 40 destructive measurements. The information on coppices' positions extracted from the aerial photographs and the topographic grid used as covariates explained a significant proportion of the millet growth variability. The use of these covariates also improved the precision of the analysis of variance of millet dry matter data by reducing the residual sum of squares by as much as 33% in the first experimental year. The study demonstrates the potential of non-destructive measurements integrated in a GIS to improve the collection and interpretation of data from field experiments.
Materialart:
Digitale Medien
URL:
http://dx.doi.org/10.1023/A:1004633313512
