Ws. Dvorak et Kd. Ross, 3-YEAR GROWTH AND STABILITY OF HONDURAN PROVENANCES AND FAMILIES OF PINUS-TECUNUMANII, Forest ecology and management, 63(1), 1994, pp. 1-11
Two orthogonal sets of five trials of P. tecunumanii were established
by the CAMCORE Cooperative at several locations in Brazil, Guatemala,
South Africa and Venezuela. Sets A and B included 46 and 35 half-sib f
amilies, respectively, from the provenances of Jocon, San Francisco, S
an Esteban and Villa Santa, Honduras. Results from assessment of 3-yea
r height indicated that trees from San Francisco, San Esteban and Vill
a Santa were equally productive across all sites and should be further
tested on a modest scale. Provenance X site interaction in both sets
was contributed primarily by the poorest source, Jocon, Of the familie
s tested, 17% were found to have b coefficients less than 0.85 or more
than 1.15 or R(2) values less than 0.90 and were classified as intera
ctive across environments. The relatively few families causing the sig
nificant interaction and the positive significant Spearman rank correl
ations (r) between most sites of close geographic proximity suggested
that genotype X site interactions were not serious. However, Spearman
rank correlations (r) between sites separated by thousands of kilomete
rs were often less than 0.30 and not significant. When selections from
one location are moved large geographic distances, they should first
be tested in trials in the new environment before being included in th
e breeding population.