A bibliographical and herbarium investigation on the pollination syndrome of Mexican columnar cacti (tribe Pachycereeae) was conducted. Most Mexican species of columnar cacti show a chiropterophilic-pollination syndrome and they flower synchronously in March to May. The floral biology, reproductive system and visitors (to both fruits and flowers) of Neobuxbaumia tetetzo, the most abundant and dominant columnar cactus of succulent forest in the Tehuacan Valley, were studied. This species reached densities of c. 1200 individuals ha−1. The bats Leptonycteris curasoae and Choeronycteris mexicana were its only pollinators, whilst a more diverse array of visitors disperse seeds. Contrary to findings for multiple pollinators of columnar cacti in extratropical deserts in North America, the relationships between N. tetetzo and nectar-feeding bats was strong and tightly coupled in Mexico.

Pollination biology, breeding system, and floral phenology of the columnar cactus Stenocereus stellatus were studied in wild, wild managed in situ and cultivated populations of central Mexico, in order to examine whether these aspects have been modified under domestication and whether they determine reproductive barriers between wild and manipulated individuals. Individuals of both wild and manipulated populations are self-incompatible, indicating that artificial selection has not modified the breeding system. Their pollination biology is also similar. Anthesis is mainly nocturnal, with a peak of nectar production between 0200 and 0400 when the stigma presents maximum turgidity. Nocturnal visitors are the effective pollinators. Nearly 75% of flowers exposed for nocturnal pollination set fruit, while none of the flowers exposed for diurnal pollination produced fruits. The bats Leptonycteris curasoae, L. nivalis, and Choeronycteris mexicana (Glossophaginae) are the most likely pollinators, and their time of foraging is synchronized with the time of nectar production and stigma receptivity in S. stellatus. Bats potentially move pollen over a considerable distance, so there is apparently no spatial isolation to prevent pollen exchange between wild and cultivated populations. Phenological studies showed that there are also no apparent temporal barriers. However, manual cross pollination failed between some domesticated and wild phenotypes, suggesting that gene flow between wild and cultivated populations might be limited by pollen incompatibility.

Summary AimThis paper examines the migration of the tropical nectarivorous bat Leptonycteris curasoae considered as a latitudinal migrant that breeds in south‐west United States and northern Mexico in spring and migrates southward during fall. We tested the hypothesis that the latitudinal migration occurs only locally given by the local availability of bat resources, leading to migratory movements in zones with seasonal scarcity of resources and to resident bat populations where resources are available throughout the year. LocalizationWe analysed the presence of L. curasoae along its distribution range in North America (between 14°N and 33°N). Study cases were also conducted in three Mexican localities: the Tehuacán Valley (17°48′–18°58′N and 96°48′–97°43′W), Sonoran Desert (28°41′N and 110°15′W), and the coast of Jalisco (19°32′N and 105°07′W). MethodsGeographic evidence for latitudinal migration of L. curasoae was analysed using 94‐year capture records housed in twenty‐two collections of North America. Records were analysed using a Geographical Information System (GIS), in which floral resources and capture records were integrated. Monthly captures in the Tehuacán Valley were conducted during three years and bats abundance and reproductive status were correlated with the phenology of bat resources. Bat captures were also conducted during two consecutive years in an extratropical desert during winter and spring, and during one spring in the coast of Jalisco. ResultsThe latitudinal migration of L. curasoae in North America only occurs at latitudes near 30°N, whereas bats may be residents at latitudes lower than 21°N. Captures were associated always to the availability of floral resources in both geographical and local scales. Main conclusionsThe existence of resident populations in the tropics with two reproductive events support the hypothesis that migration only occurs in the northern distribution limit of this nectar‐feeding bat.