Background: The breeding system of Mammillaria magnimamma was initially described as hermaphroditic. However, some individuals in a natural population at Valle del Mezquital exhibited a seemingly non-functional androecium. Objectives: To determine the breeding system of M. magnimamma through morphological characterization of floral types. Study site and years: The study was conducted at Valle del Mezquital, Mexico, during the 2019 reproductive season. Methods: We characterized floral polymorphism by means of scanning electron microscopy and morphometric analysis of 18 floral traits. Pollinator frequency, size and number of seeds, and germination rate were compared between flower types. Results: 87.85 % of the plants were hermaphrodites, and 12.15 % were females with non-functional androecium, with significantly smaller stamens, indehiscent anthers, and malformed pollen. Female flowers were shorter (14.007 ± 0.742 mm) and had smaller (1.856 ± 0.151 mm) nectar chambers than hermaphroditic flowers (15.821 ± 0.450 and 2.194 ± 0.090 mm, respectively). Although pollinators visited hermaphrodites (124 visits) about twice as often as pistillate flowers (60 visits), the latter produced more numerous ( F 1,26 = 8.320, P = 0.003) and larger ( Hc = 4.637, P = 0.031) seeds that also germinated faster ( Hc = 70.59, P < 0.0001). Conclusions: This M. magnimamma population exhibits a gynodioecious breeding system with both hermaphroditic and male sterile female plants. Females produce more, higher quality seeds, which favour their maintenance in the population.

Antecedentes: Las flores son elementos reproductivos costosos que presentan balances costo/beneficio en sus estructuras. La diversidad de éstas se ha atribuido a factores bióticos como las interacciones con sus visitadores florales y abióticos como las características del ambiente. Preguntas: ¿La morfología floral de M. geometrizans está relacionada con los tipos de vegetación? ¿La comunidad de artrópodos visitadores florales se relaciona con la morfología floral y/o el tipo de vegetación? ¿La morfología floral y la comunidad de artrópodos tendrán relación con la perturbación del ambiente? Especie de estudio: Myrtillocactus geometrizans (Mart. ex Pfeiff.) Console Sitio de estudio y fecha: Selva Baja Caducifolia y Matorral Xerófilo en las Reservas de la Biosfera Tehuacán-Cuicatlán, Puebla Barranca de Metztitlán, Hidalgo. México. Marzo 2017. Métodos: En los cuatro sitios (dos días por sitio), durante el pico de floración de M. geometrizans, se calculó el índice de disturbio, se recolectaron 30 flores (N = 120) y los artrópodos visitadores florales. Se determinaron la morfología floral y los artrópodos. Resultados: Los sitios de la Barranca de Metztitlán presentaron el mayor índice de disturbio, las flores más chicas y mayor número de estructuras florales. Los visitadores florales fueron más abundantes en los matorrales xerófilos, particularmente en el sitio más perturbado donde el grupo funcional más relevante fue el de los polinizadores. Conclusiones: Se encontraron diferencias significativas en la morfología floral de M. geometrizans relacionadas a los tipos de vegetación y ubicación geográfica. Se reportó una compensación entre tamaño y producción de estructuras florales.

Mammillaria parkinsonii and M. perbella are species endemic to the Queretano–Hidalguense Desert (QHD), a potential globose cacti Pleistocene refuge. We tested if the phylogeographic and genetic structures of these species are temporally consistent with the Pleistocene glacial refuge hypothesis. In this case, there would be high haplotype divergence associated with population contraction during the Last Glacial period (115-11.7 kya). Chloroplast DNA sequences and ISSR nuclear markers were used to estimate genetic diversity and structure, haplotype networks, and divergence times. Changes in paleodistribution were modeled using the Ecological Niche Model (MNE). cpDNA and ISSR markers revealed low-to-moderate genetic diversity in M. pakinsonii and M. perbella (π = 0.0004, 0.0044 respectively; H_S = 0.362–0.434, 0.293–0.400, respectively) and high population differentiation (G_ST = 1, 0.722, respectively; Φ_ST = 0.378, 0.3957, respectively), with no evidence of gene flow (Nm = 0, 0.29, respectively). The estimated divergence times and MNE showed that population divergence began during the Last Glacial in M. perbella and continued into the Holocene for both species. This divergence is associated with population contraction, geographic isolation, and disruption of gene flow. This joint evidence indicated that Pleistocene–Holocene glacial cycles may have driven the recent divergence of these Mammillaria species in the QHD.