Background and AimsEthnobotanical studies in Mexico have documented that Mesoamerican peoples practise systems of in situ management of wild and weedy vegetation directed to control availability of useful plants. In situ management includes let standing, encouraging growing and protection of individual plants of useful species during clearance of vegetation, which in some cases may involve artificial selection. The aim of this study was to review, complement and re-analyse information from three case studies which examined patterns of morphological, physiological and genetic effects of artificial selection in plant populations under in situ management in the region.MethodsInformation on wild and in situ managed populations of the herbaceous weedy plants Anoda cristata and Crotalaria pumila, the tree Leucaena esculenta subsp. esculenta and the columnar cacti Escontria chiotilla, Polaskia chichipe and Stenocereus stellatus from Central Mexico was re-analysed. Analyses compared morphology and frequency of morphological variants, germination patterns, and population genetics parameters between wild and managed in situ populations of the species studied. Species of columnar cacti are under different management intensities and their populations, including cultivated stands of P. chichipe and S. stellatus, were also compared between species.Key ResultsSignificant differences in morphology, germination patterns and genetic variation documented between wild, in situ managed and cultivated populations of the species studied are associated with higher frequencies of phenotypes favoured by humans in managed populations. Genetic diversity in managed populations of E. chiotilla and P. chichipe is slightly lower than in wild populations but in managed populations of S. stellatus variation was higher than in the wild. However, genetic distance between populations was generally small and influenced more by geographic distance than by management.ConclusionsArtificial selection operating on in situ managed populations of the species analysed is causing incipient domestication. This process could be acting on any of the 600–700 plant species documented to be under in situ management in Mesoamerica. In situ domestication of plants could be relevant to understand early processes of domestication and current conditions of in situ conservation of plant genetic resources.

Polaskia chichipe, a columnar cactus, is cultivated for its edible fruits in central Mexico. This study analyzed whether artificial selection has modified its reproduction patterns and caused barriers to pollen exchange between wild, managed in situ, and cultivated populations. Anthesis was diurnal (∼16 h in winter, ∼10 h in spring) as well as partly nocturnal (∼12 h in winter, ∼3 h in spring), and flowers were pollinated by bees, hummingbirds, and hawk moths. Manual cross-pollination was ∼37-49% effective in all populations. Self-pollination was ∼12% successful in the wild, but twice as successful (∼22-27%) in managed and cultivated populations. Diurnal pollination was ∼35-55% effective in winter and 100% in spring. Nocturnal pollination was successful only in winter (15%). Crosses among individuals were more effective within populations than among populations, including populations under a similar management regimen. The least successful crosses were between wild and cultivated populations. Flowers were produced in all populations from January to March, but flowering peaks differed by 1 mo among wild, managed, and cultivated populations and by 2 mo between wild and cultivated populations. The latter interrupted pollen exchange in May. Seeds from managed and cultivated populations germinated faster than those from wild individuals. Domestication has seemingly favored self-compatible P. chichipe plants with higher fruit yield, a longer period of fruit production, and faster seed germination, attributes that have resulted in partial reproductive barriers between wild and manipulated populations.

The columnar cactus Stenocereus stellatus is used in Central Mexico for its edible fruits which are harvested in wild, managed in situ and cultivated populations. Management in situ of wild populations is conducted by selectively sparing and enhancing the abundance of plants with desirable phenotypes when fields are cleared for agricultural use. Cultivation of desirable phenotypes is carried out by vegetative propagation in homegardens. Effects of human management on morphological and genetic variation of S. stellatus were analyzed by comparing morphological diversity indices (MD, based on Simpson's index) and expected (H e ) heterozygosity indices from allozyme analysis, in wild, managed in situ, and cultivated populations from La Mixteca and the Tehuaca´n Valley regions. Morphological diversity was similar among regions, but populations from the wetter La Mixteca region averaged higher genetic variation (H e = 0.279) than populations from Tehuaca´n (H e = 0.265). On average, populations manipulated by people had higher levels of variation (MD = 0.479 ± 0.012, H e = 0.289 in cultivated populations; MD = 0.461 ± 0.014, H e = 0.270 in managed in situ populations) than wild populations (MD = 0.408 ± 0.017, H e = 0.253), which is apparently due to a continual introduction and replacement of plant materials in the manipulated populations. The results illustrate that human management may not only maintain but also increase both morphological and genetic diversity of manipulated plant populations in relation to that existing in the wild. Managed in situ and cultivated populations of S. stellatus are important reservoirs of variation, and are crucial for the general maintenance of diversity in wild populations. These populations may play a principal role in designing strategies for the conservation of variation of this cactus.