Three ways a cactus occupies the landscape
Myrtillocactus schenckii, locally called garambullo, is a tree-like cactus with edible fruits. In the Puebla communities studied, it grows wild in thorn scrub, forms part of agroforestry systems and occurs in home gardens. Each situation links the plant with a different set of human actions. Fruits are gathered from wild populations. Some plants are retained when a field is prepared. A young plant can be transferred to a garden, or a section of a selected stem can be planted there. Decisions made during everyday plant use can consequently influence population composition. [Blancas et al. 2009]
Retaining plants where they already grow is a form of in situ management. Compared with planting a new individual, it involves choosing among plants already present in the landscape. The residents interviewed valued edible fruits as well as shade, living fences and the continuation of inherited practices. Selection for fruit number and size was one component of these decisions. Domestication can therefore be investigated by comparing plant traits with the reasons for retaining particular individuals. This approach connects local community knowledge with biological measurements. [Blancas et al. 2009]
Measuring the effects of plant selection
Blancas and colleagues studied three wild populations, three managed in agroforestry systems and two cultivated populations around San Luis Atolotitlán and Santiago Coatepec. The measurement design covered 240 adult plants and 26 vegetative and reproductive traits. Interviews and surveys were also conducted in 30 households in the first community and 35 in the second. Bringing these datasets together allowed stated preferences to be compared with variation observed in the field. Measurements included fruit dimensions, mass, pulp and seed characteristics, and the number of fruits per branch. [Blancas et al. 2009]
Mean fruit mass increased from approximately 0.837 g in wild populations through 0.960 g in agroforestry systems to 1.089 g under cultivation. Garden populations also produced more fruits per branch. These differences followed preferences for larger and more abundant crops. At the same time, measurements of sugar content showed no significant separation among the groups. A preference expressed by residents and a difference detected in a particular measurement are therefore separate elements of evidence. The complete set of traits formed continuous variation, with some overlap among plants from different management regimes. [Blancas et al. 2009]
More flowers and different seasonal peaks
Ortíz and colleagues extended the comparison to flower function. Near San Luis Atolotitlán they observed a wild population, plants retained in fallow fields and cacti in seven gardens. Monthly phenological observations were made for a year on ten plants in each group. Buds, open flowers and fruits were counted on twenty main branches per plant, representing four exposure directions. This sampling approach reduces the effect of selecting only the most sunlit stems and describes the season through successive reproductive stages. [Ortíz et al. 2010]
During the observed cycle, flowering lasted from early September to March. Both managed groups peaked in December, while the wild population peaked in February. Flowering periods overlapped substantially, although flower abundance was distributed differently through time. Garden plants produced markedly more flowers per branch. These differences changed the food available to visitors and the timing of the greatest pollen supply at each site. They describe a particular cycle in the studied landscape, whose course may depend on local conditions. [Ortíz et al. 2010]
Insect behaviour matters for pollination
In spring, flowers usually functioned for approximately 11 hours. They opened in the morning, presented pollen around the middle of the day and closed in the evening. Video recordings connected the timing of visits with insect movements. In cultivation, the large bee Xylocopa mexicanorum was the most frequent visitor. It landed with its body on the reproductive structures and contacted both anthers and stigma in almost all recorded visits. Pollen accumulated especially on its abdomen and legs. This contact arrangement provided a mechanical route between the site of pollen collection and the pollen-receiving surface. [Ortíz et al. 2010]
The wild population was dominated by the small stingless bee Plebeia mexicana. It frequently reached nectar by a lateral route between the tepals, making full contact with both anthers and stigma during only some visits. In managed fields, Apis mellifera and a fly of the genus Tabanus were frequent visitors. The authors assessed their effectiveness from visit frequencies, contact with floral organs and overall fruit production. The findings indicate a particularly important role for carpenter bees in gardens. Quantifying each species’ precise contribution to seed production would additionally require relating fruit development to individual controlled visits. [Ortíz et al. 2010]
Spontaneous self-pollination and cross-pollination
Breeding-system experiments used 30 flowers per treatment in each population group. Flowers were left freely accessible, isolated without further manipulation, or hand-pollinated with pollen from other individuals after their own stamens had been removed. Isolation without manipulation tested spontaneous fruit production. Cross-pollination assessed the outcome after pollen from several other plants had been supplied. Mature fruits and seeds were counted in each procedure. These outcomes describe different stages of reproductive success and are consequently informative when considered together. [Ortíz et al. 2010]
Under open pollination, 83.3% of garden flowers produced fruits, compared with 40% in the wild population and 26.7% in agroforestry systems. Following isolation without manipulation, the corresponding values were 23.3%, 3.3% and 10%. Fruits produced in gardens through spontaneous self-pollination also contained more seeds than fruits produced by that treatment in the other groups. Cross-pollination was successful in every group. The studied plants therefore combined benefits from pollen supplied by other individuals with differing capacities to reproduce using their own pollen. The greater success of the latter process in gardens is one mechanism considered in explaining their abundant fruit production. [Ortíz et al. 2010]
Selection and gene exchange act together
Choosing high-yielding plants can increase the frequency of favourable traits in a garden. At the same time, a shared landscape allows contact with wild populations. Pollen flow depends on overlapping flowering and animal movements. Seeds and plant fragments provide additional routes for transferring material between sites. The authors of both studies therefore considered domestication as a balance between human selection and processes maintaining population connectivity. Differences in flowering peaks and visitor assemblages could partially restrict pollen exchange while reproductive periods continue to overlap. [Blancas et al. 2009] [Ortíz et al. 2010]
The importance of individual mechanisms remains a hypothesis requiring the contributing factors to be separated. Field studies compared plants growing in different environments. Gardens, fields and scrub differ in conditions, flower density and animal availability. The authors associated some flowering differences with this heterogeneity, since interviews did not reveal deliberate selection for flowering time. Quantitatively separating inheritance from habitat effects would require comparisons of offspring and plants maintained under shared conditions. Morphology and pollination observations identify possible processes that such experiments could resolve more precisely. [Blancas et al. 2009] [Ortíz et al. 2010]
What a seed experiment adds
Guillén and colleagues tested another stage of the life cycle by comparing germination in four columnar cacti. Seeds of Myrtillocactus schenckii came from wild, in situ-managed and cultivated populations. After collection and storage, material was pooled into samples representing the management regimes. Germination was tested at 25°C under a twelve-hour photoperiod for 40 days. Each treatment had four replicates of 25 seeds. PEG solutions established water potentials of 0.0, −0.2 or −0.4 MPa. This provided a controlled comparison of water availability while other conditions were held constant. [Guillén et al. 2009]
Germination of Myrtillocactus schenckii was most effective at the intermediate potential of −0.2 MPa. Responses changed with water potential and sample origin, but the overall findings did not support the predicted general increase in the water requirements of seeds from cultivated populations. The sampling method matters. Pooling seeds from randomly selected plants describes an average population response that includes individuals with different characteristics. Directly testing the offspring of particular plants selected for large fruits would answer a different question. Laboratory conditions define the scope of this result, while seedling survival in the field represents a subsequent stage requiring separate observation. [Guillén et al. 2009]
Domestication that retains diversity
The studied populations show an early stage of domestication in which plant choice influences yield and the morphological composition of managed stands. Fruits, flowers, pollinators and seeds reveal different parts of this process. Greater yield can reflect flower production, the capacity for self-pollination and successful transport of pollen from other individuals. Their interaction in a garden connects human actions with animal activity. This interpretation draws on the combined picture from several studies while preserving the distinct methods and conditions of each. [Blancas et al. 2009] [Ortíz et al. 2010] [Guillén et al. 2009]
Plant variation persisted in both wild and managed sites. Traditional agroforestry systems and gardens therefore matter for conserving this cactus’s resources alongside natural vegetation. Managed plants remain part of a wider ecological system that includes pollinator feeding sites and routes for moving seeds and cuttings. The history of garambullo shows that biological effects of selection can already be visible while different forms of management coexist within a small area. Understanding that history requires knowledge of both the plants and the practices of the people who retain them. [Blancas et al. 2009] [Ortíz et al. 2010]
Traditional Management and Morphological Patterns of Myrtillocactus schenckii (Cactaceae) in the Tehuacán Valley, Central Mexico
Blancas J., Casas A., Lira R., Caballero J., Traditional Management and Morphological Patterns of Myrtillocactus schenckii (Cactaceae) in the Tehuacán Valley, Central Mexico, Economic Botany, 2009, 63(4):375-387, doi: 10.1007/s12231-009-9095-2
The study combines interviews about plant use with comparisons of wild populations, plants retained in agricultural systems and plants cultivated in gardens. Local people selected, among other traits, abundant fruit production and preserved desirable plants during land conversion. Cultivated populations differed in fruit size and number, whereas sugar content did not differ significantly among management types. The authors interpret this pattern as an early stage of domestication.
Pollination biology of Myrtillocactus schenckii (Cactaceae) in wild and managed populations of the Tehuacán Valley, México
Ortíz F., Stoner K. E., Pérez-Negrón E., Casas A., Pollination biology of Myrtillocactus schenckii (Cactaceae) in wild and managed populations of the Tehuacán Valley, México, Journal of Arid Environments, 2010, 74:897-904, doi: 10.1016/j.jaridenv.2010.01.009
The paper compares flowering, insect visits and mating systems in populations under different forms of management. Home-garden plants produced more flowers and fruits, and large Xylocopa mexicanorum bees were important flower visitors. Controlled pollination showed predominantly outcrossed reproduction with some capacity for self-pollination. The authors also described seasonal variation in flower opening duration.
Seed germination of wild, in situ-managed, and cultivated populations of columnar cacti in the Tehuacán-Cuicatlán Valley, Mexico
Guillén S., Benítez J., Martínez-Ramos M., Casas A., Seed germination of wild, in situ-managed, and cultivated populations of columnar cacti in the Tehuacán-Cuicatlán Valley, Mexico, Journal of Arid Environments, 2009, 73(4–5):407-413, doi: 10.1016/j.jaridenv.2008.12.018
The study compares germination in four cactus species under different water potentials with fixed temperature and lighting conditions. Material came from wild and human-managed populations. Myrtillocactus schenckii germinated most slowly among the species examined, and its response depended on experimental conditions. The results did not support a general increase in germinating seeds’ water requirements with increasing management intensity.