Provenance variation in Pinus caribaea, P. oocarpa and P. patula ssp. Tecunumanii

An international study of provenance variation in Pinus caribaea Morelet and P. oocarpa Schiede in Central America and the Caribbean started in the early 1960s. By the early 1970s seed had been distributed to over fifty countries for many hundreds of trials. By the late 1970s the most comprehensive of these had reached seven years, the optimum age for evaluation. A coordinated assessment programme was undertaken in a set of 29 trials in which 25 P. caribaea and 20 P. oocarpa provenances were represented. Four of the P. oocarpa sources were later recognized as P. patula Schiede andamp; Deppe ssp. tecunumanii (Equiluz andamp; Perry) Styles. Provenance representation was comprehensive for P. caribaea vars. hondurensis Barr. andamp; Golf. and caribaea, except for the omission of a few isolated populations; only one var. bahamensis Barr. andamp; Golf. provenance was included. Representation was satisfactory for P. oocarpa in Central America, but none of the populations in the central part of the range of P. patula ssp. tecunumanii in Honduras and Guatemala was sampled. New international trials of both P. caribaea var. bahamensis and P. patula ssp. tecunumanii are now in progress. Overall test site coverage was good except for P. caribaea vars. bahamensis and caribaea. Eighteen traits were assessed to describe productivity, stem quality, branching, reproduction, wood density and oleoresin composition. These are summarized in provenance-locality tables. Analysis of the data revealed statistically and operationally significant general differences between provenance regions for all three species as well as local differences between provenances within the regions. Five major groups emerged in P. caribaea, viz. var. caribaea, var. bahamensis, and inland, coastal and insular var. hondurensis. Each had productivity and silvicultural attributes that were in themselves of sufficient economic importance to influence choice for operational use in different sets of circumstances. Four provenance regions could be distinguished for P. oocarpa, the Mexican, North Guatemalan, South Guatemalan and Honduras-Nicaraguan. The Mexican type had poor productivity and undesirable silvicultural features but the differences between the other three were in traits that had little or no immediate operational significance. The four P. patula ssp. tecunumanii provenances separated on branching traits and resin composition into a group of the three Nicaraguan populations and the one from Belize. Although geno type environment interaction effects were not statistically estimated here, there were indications that they exist at the provenance region level for the three varieties of P. caribaea particularly where wind and biotic effects are significant factors in the environment. These trials have provided data that can be utilized not only to make immediate operation-al gains through provenance selection, but also to plan breeding strategies to enhance populations and to create synthetic varieties in a systematic genetic development for each species.