Within-stand spatial distribution of tree mortality caused by the Douglas-fir beetle (Coleoptera : Scolytidae)

Citation
Jf. Negron et al., Within-stand spatial distribution of tree mortality caused by the Douglas-fir beetle (Coleoptera : Scolytidae), ENV ENTOMOL, 30(2), 2001, pp. 215-224
Citations number
48
Categorie Soggetti
Entomology/Pest Control
Journal title
ENVIRONMENTAL ENTOMOLOGY
ISSN journal
0046225X → ACNP
Volume
30
Issue
2
Year of publication
2001
Pages
215 - 224
Database
ISI
SICI code
0046-225X(200104)30:2<215:WSDOTM>2.0.ZU;2-N
Abstract
The Douglas-fir beetle, Dendroctonus pseudotsugae Hopkins, causes considera ble mortality in Douglas-fir, Pseudotsuga menziesii (Mirb) France, forests. Within-stand distribution of mortality was examined in affected stands usi ng geostatistical techniques. A 10 x 10 m grid was established in two 4-ha study sites. Live and beetle-killed host basal area was measured at each nu de. In a 16-ha stand, a variable-resolution grid was established and the sa me information collected. The relationship between Douglas-fir basal area a nd Douglas-fir basal area killed war examined using non-spatially explicit and spatially explicit Linear regression models. A positive linear relation ship was observed between the variables. Significant spatially explicit mod els suggest that the relationship is also true at fine scales. Relative var iograms were constructed for Douglas-BI basal area before and after the Dou glas-fir beetle outbreaks. For the 4-ha sites, increased spatial dependency in the distribution of Douglas-fir basal area was observed as a result of the Douglas-fir beetle outbreak. For the 16-ha site, kriging was used to es timate live Douglas-fir basal area before and after the outbreak to a 10-m resolution and the stand rated for potential mortality illustrating the pot ential applicability of geostatistical techniques to rating a stand far pot ential mortality. Cross-validation analysis indicated that although the pot ential exists for large estimation errors, the majority of the estimates we re within acceptable ranges. The study suggests that geostatistical approac hes map be suitable to extend our understanding bark beetle ecology and imp roving the application of extent of mortality models.