The path of pollen in Ariocarpus. Partial self-incompatibility and seed set

Flower structure favours pollen transfer between plants

The flowers of the seven Ariocarpus species studied were bisexual and opened during the day. The pistil extended above the stamens, placing the stigma and anthers at different heights. This spatial separation is called herkogamy. In a comparison of 216 flowers, the mean difference between pistil and stamen length ranged from 3,59 mm in A. kotschoubeyanus to 5,02 mm in A. agavioides (A. agavoides in the publication). Each species was represented by one population. [Martínez-Peralta et al. 2014a, pp. 665–670, Table 1]

Pollen release often began before the stigma became fully receptive, but the male and female phases overlapped. Herkogamy and incomplete protandry can alter an insect's route and the timing of pollen availability. The behaviour of self-pollen after contact with the stigma was assessed directly in a pollen-tube experiment. It also included A. scaphirostrus (A. scaphirostris in the publication). [Martínez-Peralta et al. 2014a, pp. 668–672; pp. 532–536][Martínez-Peralta et al. 2014b, pp. 668–672; pp. 532–536]

Microscopy traced five sections of the route

The self-incompatibility study included 562 pistils from natural populations, collected during the 2005–2010 seasons. Natural pollination, hand self-pollination and hand cross-pollination were compared for each species. In the cross-pollination treatment, anthers were removed and pollen from at least five plants was mixed. Hand pollination was performed on both days of each flower's life between 13:00 and 15:00. The flowers remained covered from the bud stage until flowering ended. [Martínez-Peralta et al. 2014b, pp. 532–533]

The pistils were fixed 48 hours after the last pollination and stained with aniline blue, which makes pollen tubes visible under fluorescence microscopy. Their presence was recorded on the stigma, in three successive zones of the style and in the ovary. Squash preparations established whether tubes were present or absent without counting every tube precisely. The percentage of pistils with a tube in the ovary therefore measures completion of the route by at least one tube. [Martínez-Peralta et al. 2014b, pp. 532–534, Fig. 1 and 3]

Self-pollen was often arrested in the style

Pollen adhered to the stigma in most flowers examined, and germination also occurred after self-pollination. Differences increased during subsequent growth. After self-pollen was used, tubes reached the ovary much less often. Arrest in the style is consistent with gametophytic self-incompatibility, in which recognition during growth of the male gametophyte determines compatibility. The study established pollen-tube behaviour. It did not identify compatibility genes at the molecular level. [Martínez-Peralta et al. 2014b, pp. 533–536, Fig. 3–4]

Self-pollen was strongly inhibited, but the degree of inhibition varied between species and samples. In six species, self-pollen tubes sometimes reached the ovary. In A. bravoanus, this occurred in one of only three pistils in this treatment. In A. scaphirostrus, tubes were not found in any of the 37 pistils, although a separate trial produced a single fruit after self-pollination. These findings support predominantly outcrossing reproduction and varying degrees of partial self-incompatibility. [Martínez-Peralta et al. 2014b, pp. 534–536, Table 1]

From a pollen tube in the ovary to a mature fruit

Pistils used for microscopy were fixed, while fruits were collected from a different group of flowers five to six months after pollination. In many combinations, fruit set was lower than the frequency of pollen tubes in ovaries suggested. The arrival of a tube in this zone does not determine how many ovules will be fertilized or how many developing seeds will reach maturity. [Martínez-Peralta et al. 2014b, pp. 532–535, Table 1]

The authors considered several further constraints, including too few successful pollinations, early loss of offspring after self-pollination and the plant's available resources for fruit development. The individual causes were not separated experimentally. The comparison identifies where self-pollen is inhibited and shows that further processes come into play at the fruit stage. [Martínez-Peralta et al. 2014b, pp. 535–537]

Adding pollen increased seed number while the proportion of fruits remained similar

In a population of A. kotschoubeyanus near Tula, flowers naturally visited by insects were compared with flowers additionally given a pollen mixture from ten plants in other patches of the site. Pollen was added on both days of flowering. After a nocturnal herbivore damaged some flowers, the final samples comprised 17 control flowers and 19 given additional pollen. Each came from a different plant. [Martínez-Peralta and Mandujano 2016, pp. 3–5]

They produced 14 and 16 fruits respectively, representing about 82% and 84% of the flowers. The difference in fruit set was not significant. However, naturally pollinated fruits contained a mean of 33,57 ± 12,85 seeds, compared with 43,43 ± 8,38 after additional pollen. The reported variation is the standard deviation. Mean seed number was related to a mean of 45,8 ovules, determined from a separate set of 36 flowers. Seed set calculated in this way rose from 0,73 to 0,95, an increase of 22 percentage points. [Martínez-Peralta and Mandujano 2016, pp. 4–5, Table 1]

In this trial, additional pollen improved the filling of fruits with seeds. The treatment increased pollen supply and used multiple donors, so the result does not separate the effect of pollen quantity from that of compatibility. Mass flowering and frequent fruit set can coexist with this kind of limitation on seed set. Counting fruits and counting seeds provide complementary answers here. [Martínez-Peralta and Mandujano 2016, pp. 4–7]

Phenology and insects define the local context

In A. fissuratus at Cuatrociénegas, peak flowering lasted about five days in 2005 and three in 2006. One-day flowers predominated in the first season and two-day flowers in the second. Bees touched the stigma before reaching the anthers, creating an opportunity to deposit pollen carried from a previous visit. A small series of flowers isolated from insects produced no fruits through spontaneous self-pollination. [Martínez-Peralta and Mandujano 2011, pp. 148–151]

In 2006, 45 flowers on 30 plants were also observed to assess flower consumption by darkling beetles. Fourteen flowers were eaten completely and seven partially. Removal of the whole flower prevented it from setting fruit. Partial damage produced a varied group depending on which organs remained. Flower consumption, pollen transfer and pollen compatibility thus concern different points in the same reproductive cycle. [Martínez-Peralta and Mandujano 2011, pp. 149–152]

The visitor assemblage also varied between populations. Honeybees dominated observations of A. fissuratus, while on A. kotschoubeyanus in Tolimán, most visits were recorded for native bees of the genera Augochlora and Agapostemon. At the latter site, some bee taxa were shared with Neolloydia conoidea, which flowered earlier in the year. The course of this comparison is described in the account of shared bees and flowering sequence. [Martínez-Peralta and Mandujano 2011, pp. 150–153; pp. 29–35, Table 2][Arroyo-Pérez et al. 2021, pp. 150–153; pp. 29–35, Table 2]