The known vertebrate fauna of the barrier island of Tamaulipas consists of one species of tortoise, two species of lizards, at least one (unidentified) species of snake, 49 species of birds (48 recorded by us and the Semipalmated Sandpiper), and 12 species of mammals. This is clearly a depauperate fauna, such as is characteristic of islands generally, and indicates that the peninsular nature of the northern part of the barrier island is of relatively small consequence in determining presence or absence of species. It is likely that the restricted environmental spectrum is much more important in this regard than is the fact of semi-isolation.

Of the 49 species of birds, 10 are known to breed on the island and an additional 21 are suspected of breeding either on the island or on small islets in the adjacent Laguna Madre de Tamaulipas. Eleven species occur on the island as nonbreeding summer residents, about which we will have more to say below. Four species have been recorded on the island in summer but breed elsewhere, that is to say, they only wander over the island (Man-o'-war Bird, Turkey Vulture, etc.). Two species are known only as migrants, and the status of one, the Sora Rail, is uncertain. The number of migrant species doubtless will be greatly increased by field work at those times when birds migrate.

The avifauna is not depauperate owing to the exclusion of any one of the three major zoogeographic stocks thought to be important in the development of the present North American avifauna (Mayr, 1946). If we examine the breeding passerine birds of the barrier island and the breeding passerine assemblage at the same latitude in lowland Sonora (Mayr, loc. cit.) as to their ultimate evolutionary sources, we find that for both places somewhat more than half the birds have developed from indigenous, North American stocks, about one-third have been derived from South American stocks, and one-fifth to one-eighth are from Eurasian stocks. It is most unlikely that such close correspondence in relative composition of the two avifaunas would occur by chance. Thus, we can only conclude that each of the historical avian stocks is proportionately restricted in numbers on the barrier island.

Faunistically, the barrier island resembles Padre and Mustang islands and the adjacent mainland of Tamaulipas and southern Texas, reflecting the relative uniformity of environment in this region. It is apparent that there is a faunal "break" or region of transition in the vicinity of Tampico, in extreme southeastern Tamaulipas. On the coastal plain, many tropical species and subspecies occurring in Veracruz are found north to Tampico but fail to extend farther northward to the barrier island of northeastern Tamaulipas. Axtell and Wasserman (1953:4-5), have already commented on this situation, mentioning a number of snakes and lizards that have differentiated subspecifically on opposing sides of the Tampican region. They also note that large numbers of the lowland Neotropical floral and faunal elements reach their northern limits of distribution within the zone of transition around Tampico, and, also, many Nearctic elements find their southern distributional limits there.

Our small samples of birds and reptiles from the island show no detectable morphological differentiation from adjacent populations. However, several of the mammals are moderately-well differentiated, but the patterns and degrees of geographic variation are such that we can only speculate on the historical derivation of the insular populations. Lepus californicus curti is presently known only from the barrier island of Tamaulipas, but Hall (1951:43) has suggested that it may also occur on the adjacent mainland. A resemblance between individuals of this subspecies and specimens of L. c. merriami from Padre Island in smallness of the tympanic bullae is regarded, probably correctly, by Hall (1951:44) as independent development—that is, parallel adaptation to similar environmental conditions reaching fullest expression on the barrier island of Tamaulipas. As is also true with Geomys personatus and Neotoma micropus, the barrier island population of Lepus californicus shows relationships with animals from Texas and northern Tamaulipas (L. c. merriami) and no connection with (resemblance to) animals from the south (L. c. altamirae, known only from the type locality at Altamira, near Tampico).

In color and cranial proportions, Dipodomys ordii parvabullatus of the barrier island is closer to D. o. compactus of Padre Island than to D. o. sennetti of southern Texas and the Tamaulipan mainland. But, D. o. parvabullatus resembles D. o. sennetti in external measurements (Hall, 1951:39). Possibly D. o. parvabullatus and D. o. compactus are phylogentically closer to one another than is either to D. o. sennetti. It is also possible that each evolved independently from a mainland stock represented today by D. o. sennetti; the resemblance of the two insular populations would thus be a matter of convergence in response to like environmental conditions.

Sigmodon hispidus solus is an insular differentiate that probably reached the barrier island from the adjacent mainland of Tamaulipas, where its apparent closest relative, as judged by morphological similarity, now occurs.

Nonbreeding shorebirds in summer south of breeding ranges.—Certain aspects of this subject have already been discussed by Eisenmann (1951). As he notes, the phenomenon is more regular and widespread than generally has been appreciated. The old idea, that such oversummering individuals were "abnormal" or "senile," is totally inadequate, especially in view of the frequently large numbers of individuals involved.

Eisenmann's suggestion that nonbreeders are immature is probably valid, and it is supported by Pitelka's examination of dowitchers (1950:28, 51). For gulls, which can be aged by characters of plumage, there is no question that most nonbreeders are immature. Unfortunately, there are few criteria for determination of age in charadriiform birds.

With the possible exception of a specimen of Limosa fedoa, none of the presumed nonbreeding, oversummering shorebirds collected by us showed gonadal enlargement above expected minimal sizes for the species. Even so, the season was late at the time when we were on the island and most of the birds were molting; it is possible their gonads had been enlarged earlier in the season. Behle and Selander (1953) and Johnston (1956) have shown that nonbreeding first-, second-, and third-year California Gulls (Larus californicus) undergo gonadal enlargement in summer. Additionally, nonbreeding first-year males of certain passerine species (for example, the Brown Jay, Psilorhinus morio; Selander, 1959) are known to experience partial gonadal recrudescence in summer. It would be useful, and would facilitate discussion, to have data on gonadal condition of oversummering birds; any functional enlargement would be worth documenting.

Some species, notably the Semipalmated Sandpiper, Semipalmated Plover, and Black Tern, oversummer as nonbreeders in such large numbers that it is obvious that a significant fraction of the total population of the species does not breed in any one year. This raises questions concerning the possible ecologic situations that would select for delay in time of recruitment of young birds into the breeding segment of the population, assuming that nonbreeders are immature birds. Delay in maturation, or slow rates of maturation, may show general relationship to paucity of sites of breeding, as Orians (1961:308) suggests, but the shorebirds with which we are dealing breed in regions or in habitat-types not characteristically imposing general restriction on sites of nesting; more than one answer is necessary for the question even at this level. Data on age and numbers of nonbreeders, as well as on the ecology of breeding populations, are critical and are badly needed for most species.

In any event, species for which we have data demonstrating that they regularly oversummer south of their breeding ranges are probably adapted to having a part of their populations refrain from breeding each year. Whether this phenomenon can be explained solely in terms of selection at the level of individual birds (Lack, 1954) or involves selection of an adaptive response of the population as a whole (Wynne-Edwards, 1955; see also Taylor, 1961, concerning Rattus) is a problem that cannot be resolved at this time. We may note that the species involved ordinarily breed in arctic and subarctic regions, and it would seem advantageous (as set forth below) for nonbreeders to remain well south of such high latitudes. The numbers of oversummering individuals may fluctuate with over-all population density, possibly as a result of crude density, but possibly also as a result of emigration of individuals in excess of optimal density on breeding grounds (see Wynne-Edwards, 1959). One aspect of this phenomenon not explicitly discussed by Wynne-Edwards is the possibility that some individuals never move north to breeding grounds at all, perhaps as a result of a behavioral character genetically-grounded and mediated by delayed maturation of the neurohumoral "clock." This certainly would be an economical means by which population numbers could be regulated, for there would be a saving of energy in that some individuals not only would not move north, but also would not participate in the behavioral interactions involved in territorial spacing. Occurrence of these birds throughout southern North America, Middle America, and northern South America may thus reasonably be understood.