THE CENTRAL AMERICAN SPECIES OF QUERCUS 5 



In Central America both types of distribution are noted. In wide 

 areas of similar and closely associated or contiguous habitats, as, for 

 instance, the less rugged eastern watershed, very widespread species 

 are encountered. These are for the most part species of tropical rain 

 forests, especially toward the upper limits of the tropical zone. 

 Quercus insignis and Q. oleoides are examples that range variously 

 from Tamaulipas and Veracruz in Mexico to Honduras and Costa 

 Rica. The more rugged western watershed bears some widespread 

 species and some rather restricted ones. The extremely high peaks, 

 on the other hand, yield a number of endemics such as Q. costaricensis , 

 Q. irazuensis, and Q. pacayana. Although it is obvious that the basic 

 cause of this endemism is isolation of habitats, it can only be con- 

 jectured if the lack of endemism is brought about by hybridization 

 and consequent swamping of species or by mere failure to evolve due 

 to a freely interbreeding population. At least one case (that involv- 

 ing Q. peduncularis and its variety sublanosa) closely resembles hy- 

 bridization and partial swamping of that form here regarded as the 

 variety. If that is true it represents the only known case of hybridity 

 in tropical American oaks. 



Any species of Quercus is capable of surprising variation. Sometimes 

 the variations involved are numerously represented, but frequently 

 they occur as one or a few individuals. If by chance the variant 

 resembles another species and this therefore makes its identification 

 difficult, it is very apt to be termed a hybrid. Such a practice on the 

 part of taxonomists identifying specimens to be placed in herbaria 

 is highly confusing and cannot be discouraged too emphatically. To 

 state that a puzzling form is a cross between two species which it re- 

 sembles is admittedly a much easier disposition of a troublesome 

 problem than painstakingly searching out its true identity. The 

 cases of some herbaria are filled with specimens bearing the bar 

 sinister, which, were they only correctly named, would go far toward 

 presenting the student with a true picture of the range of variation 

 to be expected within various species. 



That the species concept here employed differs from that of Trelease 

 is evidenced principally by the number of species placed in synonymy 

 here. This difference of opinion seems to arise from the admission 

 of greater intraspecific variation in the present work, and this in 

 turn results from the availability of a greater quantity of specimens 

 of many species than was formerly to be had. This wealth of ma- 

 terial, for which Paul C. Standley is in large part responsible, has 

 revealed the unreliable nature of many characters on which species 

 have been based. There are still a few species that might conceivably 

 break down should additional specimens appear. 



In short, a very conservative attitude has been maintained in the 

 delimitation of species. Although this has largely simplified the 

 task of constructing a key to the species, the greater variability 

 admitted has in some cases made it even more difficult. Furthermore, 

 this conservative viewpoint has taken care of only those difficulties 

 that arise from species being based upon too ephemeral characters. 

 It has not eliminated those utterly different (and often only remotely 

 related) species that are superficially so like one another as to make 

 any distinction by key or description extremely difficult. It is hoped 

 that the illustrations will supplement the key and descriptions 

 sufficiently to deal with such problematic distinctions. 



