Three sexual forms of cardón and two ways of visiting a flower

Recognizing floral function

On the Sonoran coast, tall cardón plants develop large flowers that open after sunset. A pistil and numerous anthers are visible inside. This structural similarity long encouraged the interpretation that all plants were hermaphroditic. Theodore Fleming and colleagues instead demonstrated three functional forms. Some individuals produce both seeds and pollen, some contribute pollen alone, and others produce seeds only after receiving pollen from another plant. Their coexistence within a population is called trioecy. [Fleming et al. 1994]

Female flowers retain anthers, but these are smaller and produce no pollen of their own. Male flowers have a visible pistil, while their ovaries generally lack normally developed ovules. Determining sex therefore required dissection and examination of the floral interior. The researchers also encountered a few individuals without functioning male or female reproduction. These rare sterile plants do not constitute a fourth successfully reproducing sex. They illustrate the distinction between a retained structure and its actual reproductive function. [Fleming et al. 1994]

Plants at two Sonoran sites

Fieldwork took place during spring from 1989 to 1991 near Bahía Kino. The main site covered about 30 ha and eventually included 143 marked plants. A second site lay within a dense cardón stand. Sex proportions differed between the localities. Females were the largest group at the main site in 1991, whereas hermaphrodites were most frequent in the 1990 sample from the second stand. These results describe particular field samples and do not establish a single sex ratio for the entire species. [Fleming et al. 1994]

Reproduction was compared at several levels. The researchers counted flowers opening each night and over the season, measured nectar, quantified pollen production and followed fruits. Males produced more flowers, increasing their seasonal pollen contribution. Female plants contributed through seeds. A similar quantity of pollen in a single male and hermaphroditic flower therefore did not mean that whole plants made equal contributions. Differences emerged after flower number and the duration of flowering were taken into account. [Fleming et al. 1994]

Self-compatibility and access to pollen

Hand-pollination experiments tested what limited fruit production. In the 1990 experiment, female fruit set increased from 35% under open pollination to 73% after pollen was supplied by hand. The comparison involved 55 and 40 flowers respectively. A corresponding increase was not demonstrated in hermaphrodites. The difference indicates that pollen delivery more strongly limited seed production by unisexual female plants at this site. Reproductive performance therefore depended on the combination of plant sexual form and pollination conditions. [Fleming et al. 1994]

Self-compatible hermaphrodites could use their own pollen. However, bagged flowers left without manipulation produced fewer fruits than flowers accessible to visitors. An animal can move pollen within one plant and facilitate contact with a stigma. Self-compatibility therefore describes the capacity for fertilization by self pollen, rather than complete independence from visits. Evaluating the benefits of selfing also requires following offspring. Measurements of fruits and seeds cover only part of their subsequent survival and reproductive performance. [Fleming et al. 1994]

The geography of chromosome numbers

A later study by Carina Gutiérrez-Flores and colleagues added chromosome information to this picture. The populations examined on Catalana and Cerralvo Islands had 2n = 22. Plants with 2n = 44 occurred in the other populations included in the new chromosome counts. These are diploids and tetraploids, possessing two and four basic chromosome sets respectively. The diploid island populations were hermaphroditic, whereas tetraploid populations represented different systems involving unisexual individuals. Trioecy is therefore one regional reproductive arrangement within cardón. [Gutiérrez-Flores et al. 2018]

Combining ploidy, geography and floral structure led the authors to propose hypotheses about diverging evolutionary lineages. The association of these differences does not by itself establish how the entire sexual system originated. The paper identifies a need for crossing experiments between cytotypes, groups with different chromosome numbers. Such experiments would measure compatibility and offspring fertility. POWO accepts the name P. pringlei, and the described cytotypes represent variation within that species. [Gutiérrez-Flores et al. 2018] [2-WWW]

Two bats at the same flower

In Baja California Sur, Winifred Frick and colleagues compared two nocturnal cardón visitors. Leptonycteris yerbabuenae specializes in nectar feeding. It has an elongated face and tongue and can hover at flowers. Antrozous pallidus mainly eats large arthropods but also uses nectar. It lands on a flower and inserts its face and the front of its body. This difference in access to the floral reward prompted a question about how much pollen each animal transfers to a stigma during a single contact. [Frick et al. 2013]

Female flowers were selected because they produce no pollen of their own. Grains found on their stigmas could consequently be attributed to the animal visit. Observers and cameras recorded the first visit, after which the stigma was preserved for microscopic analysis. In total, 89 stigmas were collected, including 56 after A. pallidus visits and 33 after L. yerbabuenae visits. This experimental design connected the behaviour of a particular visitor with a measurable pollen transfer outcome, retaining the individual flower as the observation unit. [Frick et al. 2013]

Effectiveness per visit

A. pallidus deposited more pollen in one visit than the nectar specialist. The estimated ratio was 8.1, with a broad 95% confidence interval from 2.9 to 22.9. This result describes grains on a stigma after an observed contact. The difference was associated with animal behaviour. Landing and inserting the body deeply into a flower increased contact with floral structures. The authors also considered a role for grooming, but did not test that mechanism in a separate experiment. [Frick et al. 2013]

High effectiveness per contact is only one component of pollination service. A rare visitor may deliver much pollen, while a frequent visitor compensates for smaller loads through repeated visits. To incorporate this second dimension, the researchers used earlier observations of 731 flowers on 143 cacti at fourteen sites. They combined visitation frequency with the distribution of pollen loads deposited by each bat. This joined two datasets collected in different years to model effectiveness over a whole night. [Frick et al. 2013]

Seasonal balance of sexual functions

The persistence of three sexual forms can be analysed through the balance over an entire season. Fleming and colleagues estimated that females produced about 1.6 times as many seeds as the corresponding female function of hermaphrodites. The pollen contribution of males was about 1.5 times that of the male function in hermaphrodites. These comparisons concern separate components of reproduction. They cannot be added together as a simple measure of whole-plant success, because seeds and pollen represent different routes of genetic transmission. [Fleming et al. 1994]

The observed differences were insufficient to explain trioecy using the genetic models considered by the authors. Assumptions about sex inheritance, selfing frequency and inbreeding costs all mattered. Estimating pollen production also does not establish what fraction of grains actually fertilizes ovules on other plants. Models help identify missing links in a mechanism. Here, further interpretation required knowledge of both the genetics of sexual forms and spatial and seasonal variation in pollinator activity. [Fleming et al. 1994]

The timing of a visit also mattered in the bat observations. Frick and colleagues found different patterns of pollen delivery as the night progressed. In A. pallidus delivery initially increased and then levelled off. In L. yerbabuenae the pollen deposited in a single visit declined with time after sunset. Pollen transport is therefore dynamic. Even the same animal species can contribute differently at the beginning of the night and after repeated flower visits. [Frick et al. 2013]

From a stigma to offspring

The relative importance of the two bat species varied among sites. A. pallidus visited less frequently, but a large pollen load could partly compensate for fewer contacts. At many sites L. yerbabuenae visited often enough to perform an important role as well. The assessment depended on the local combination of these quantities. A single-visit result therefore does not identify one universally best cardón pollinator. It shows how different animal behaviours can support effective pollen transfer in the same plant species. [Frick et al. 2013]

Delivered pollen must subsequently germinate, produce pollen tubes and achieve fertilization, while the plant must sustain developing fruits. Frick and colleagues did not directly measure differences in final fruit and seed numbers following visits by the two bats. Their study describes an important step towards reproduction. Together with Fleming’s experiments, it also shows why the value of visits depends on a cardón plant’s sexual form. The same flowering landscape connects plants with different access to self pollen and animals that transport it in different ways. [Frick et al. 2013] [Fleming et al. 1994]