DIGITALNA ARHIVA ŠUMARSKOG LISTA
prilagođeno pretraživanje po punom tekstu



ŠUMARSKI LIST 11-12/2011 str. 46     <-- 46 -->        PDF

N. Šprem, M. Piria, H. Novosel, T. Florijančić, B. Antunović, T. Treer: MORPHOLOGICALVARIABILITY... Šumarski list br. 11–12, CXXXV (2011), 575-583


Figure 2 RDA ordination of 11 sampling sites and 16 morphological traits a) all age classes, b) juvenile c) yearling d) adults


Slika 2. RDA koordinate 11 lokacija i 16 morfoloških svojstava a) sve dobne skupine, b) mladi c) jednogodišnjaci d) odrasli


and adults showed similar results at the most locations
(Fig 2b, 2c, 2d). Results of cluster analysis usingTREE
procedure indicate separation on the two subpopulations
and suggesting the existence of morphological differences
(Fig 3).The first canonical variable accounts for
84.8, 76.5 and 84.2 percent of the total variance for the
three ages classes respectively, but is unevenly correlated
with the original variables. At first and second
canonical variables for the young shows the highest correlation
with CS, for sub-adults HH and for adults EW,
TLand CC.The discriminant analysis results in the correct
classification of the data in the groups shows that
Central and East populations are correctly classified and
a higher percentage of misclassification can be observed
forWest population. This result is not easily explainable
but the effect of hybridization can be supposed.