Mammilloydia candida. Scattered habitats and genetic diversity

From a historical name to a verified site

The species name accepted by POWO is Mammilloydia candida. The publication by Solórzano and co-authors appeared under the historically common name Mammillaria candida. The difference concerns assignment to a genus. Both names refer to the same taxon. The name used in the title allows the publication to be located under the older classification. [2-WWW][Solórzano et al. 2026, title and scope of the study]

The analysis was based on four expeditions in 2023–2025 and inspections of 62 sites in nine Mexican states. From 690 initial GBIF records, duplicates, locations of insufficient accuracy and information questioned in the field were removed, among other exclusions. The final dataset contained 254 records from Coahuila, Nuevo León, San Luis Potosí and Tamaulipas. Individual records document occurrence at a specified location. Plant numbers, occupied patch size and identification quality are separate information. [Solórzano et al. 2026, pp. 2–6]

Verification had a significant taxonomic dimension. At some historical localities, other cacti with a similar appearance were found, including Mammillaria muehlenpfordtii and M. geminispina in Guanajuato. Some sites in Querétaro had been occupied by buildings and farms. Such checking allows misidentification, habitat transformation and currently confirmed species presence to be distinguished. The authors also identify the need for further taxonomic revision. [Solórzano et al. 2026, pp. 5–6, 13–15]

Patches on hills and shelter from neighbouring plants

Confirmed occurrence sites lay at elevations of approximately 1140–2160 m above sea level. Plants grew at mountain bases, on slopes and hills, often beneath low shrubs or between rosettes of Agave, Hechtia and Nolina. The authors also recorded nearby Yucca and Opuntia. In cliff crevices, cacti occupied sites containing soil, while occurrence on open surfaces was less frequent. [Solórzano et al. 2026, pp. 5–6, 13–15]

At some sites, plants clustered in areas on the order of one hundred by one hundred metres, with large gaps between successive clusters. Some sites in Coahuila contained only three to nine plants. This produces a mosaic of habitats differing in size and history. Proximity to shrubs and rosette plants is part of the description of occupied microhabitats. [Solórzano et al. 2026, pp. 5–6, 13–15]

What the habitat model represents

The authors compared MaxEnt models using climate, soil and both variable sets. Climate data came from WorldClim 2.1, and soil properties from SoilGrids 2.0. Calculations were performed in cells of approximately 1 km². Soil data concerned a depth of 5–15 cm. The model assigns map cells a degree of similarity to the environmental conditions of known sites. [Solórzano et al. 2026, pp. 2–4]

The soil model gave the most extensive projection, also including areas without confirmed occurrence. The authors prefer the combined model for its agreement with field observations and pROC result. Individual assessment indices describe different properties of the fit. In the same table, the climate model has a lower AICc and higher TSS. Selection of the combined model is therefore an assessment bringing together several lines of evidence. [Solórzano et al. 2026, pp. 7–8, Table 2]

The largest contributions to the combined model came from isothermality, approximately 23,5%, soil cation exchange capacity, 19,7%, and precipitation seasonality, 17,5%. Cation exchange capacity describes the soil's ability to retain and exchange positively charged ions. The spatial association of this property with known sites indicates the importance of soil conditions in prediction. The contributions describe the importance of individual variables in spatial prediction. [Solórzano et al. 2026, pp. 8–9]

Two views of genetic variation

The plastid fragment ycf1–ycf2 was sequenced in 20 plants from four sites, five from each. Two haplotypes were detected. One occurred in both sampled localities in San Luis Potosí, and the other in the sampled sites in Nuevo León and Tamaulipas. This is low variation in a specific chloroplast fragment forming part of the inherited plastid genome. [Solórzano et al. 2026, pp. 3–4, 10–11]

Five nuclear microsatellite markers were examined in 95 individuals from eight sites. Microsatellites are short repeated sequences whose repeat number may differ between individuals. In this sample, the authors distinguished 138 alleles. Marker structure separated groups from San Luis Potosí, Nuevo León and Tamaulipas, and 88,8% of variation in the AMOVA analysis used was attributable to differences within populations. The result therefore combines high local variation with recognisable regional differentiation. [Solórzano et al. 2026, pp. 10–12]

The plastid and microsatellite pictures are based on different markers and different samples of individuals. The Coahuila sample for microsatellites contained only two plants, while better-represented sites included up to twenty individuals examined. Conclusions about this part of the range require additional sampling. Regional genetic groups can be considered as evidence for selecting conservation units while retaining a single accepted species. [Solórzano et al. 2026, pp. 10–12, 15–16]

A significant deficit of heterozygotes, individuals with different alleles at a given marker, was detected in some populations. The authors consider mating between related plants in small clusters and limited pollen transport. These are interpretations of marker structure. Effective pollinator species, pollen transport distances and the contribution of individual animals to seed dispersal require direct observations and experiments. [Solórzano et al. 2026, pp. 12, 15–16]

Conserving sites and regional diversity

The boundaries of the six protected areas considered encompassed 1,45% of the area indicated by the combined model. This value concerns potential habitat on the map. None of the 254 verified location records fell within those boundaries. Combining the two pieces of information reveals the gap between the existing protection network and the distribution of documented plants. [Solórzano et al. 2026, pp. 9–10, Fig. 5]

The authors documented transformations of sites in Coahuila, including industrial development, a farm and a scrapyard. They propose conserving several regional components of genetic diversity and extending sampling and taxonomic work. Protecting an occurrence site also maintains the germination environment, neighbouring vegetation and the opportunity for reproduction. [Solórzano et al. 2026, pp. 13–16]

Genus and species profiles