Reproductive ecology of the prickly pear Opuntia atropes, a native plant subject to management

Abstract

Species of Opuntia (prickly pears), widely propagated for human use, are present in many agroforestry systems. Like other perennials, they are subject to environmental pressures that influence their reproductive ecology and reproductive output. Here, we studied floral biology, breeding system, and flower-bee interactions of Opuntia atropes, which is subject to management, to better understand its reproductive ecology. We documented floral phenology, morphology, and mating system by observing live flowers, measuring fixed flowers, and performing pollination treatments. We recorded frequency, behavior, and floral visitors’ handling time to determine the most effective pollinator. Opuntia atropes has large, herkogamous flowers and high P/O ratios. These traits promote outcrossing, which was supported by the results of pollination treatments. Fruit weight and seed set exhibited moderate pollen limitation. Floral visitors comprised nine bee species. Diadasia sp. was the most effective pollinator according to frequency and behavior. Opuntia atropes shows preferential outcrossing through pollination by a specialist bee despite being in a modified habitat. Pollen limitation and variation in reproductive success may lead to mating system evolution in O. atropes. The observed bee diversity suggests that O. atropes is an important source of floral resources in the dry season in semi-arid environments under moderate human disturbance.

Keywords:
Opuntia; native bees; effective pollinator; bee diversity; floral biology; cactus; tropical dry forest

Introduction

In the face of global pollinator decline (Potts et al. 2010), maintaining native flora is a key factor for safeguarding the diversity of pollinators in natural and semi-natural habitats. Native flora provides high-quality resources for pollinators; in addition to assuring pollinator diversity, this also maintains the interactions between plants and their pollinators (Carman & Jenkins 2016). An important plant element in the drylands of North America is the genus Opuntia-the prickly pears. Species of Opuntia may grow in several types of vegetation, with varying degrees of dominance. Vegetation that is dominated by one or several species of Opuntia is commonly called nopalera (Rzedowski 1978; Rebman & Pinkava 2001). Nopaleras provide food and habitat resources for vertebrates, such as rodents, birds, and other mammals (González-Espinosa & Quintana-Ascencio 1986). Although less documented, nopaleras also harbor high insect diversity. For example, prickly pear crops have been shown to have bee and ant diversity that is on par with natural vegetation (Gómez-Otamendi et al. 2018; Ávila-Gómez et al. 2019), suggesting that Opuntia plants provide high quality resources such as food and habitat.

Several species of the genus Opuntia also represent an important food resource for humans, since their mature fruits and their shoots, called cladodes, are edible (Pimienta-Barrios 1990; Fernández-Montes et al. 2000). Opuntia species are also used as fodder for livestock and as living fences. Due to these multiple uses of Opuntia, the species of this genus have been subject to management for centuries (Bravo-Hollis 1978), and some species are completely domesticated. The most important domesticated species is O. ficus-indica, cultivated in several countries worldwide (Griffith 2004; Reyes-Agüero et al. 2005b). Other species have been subject to a less intense selection since plants are tolerated within agroforestry systems rather than actively cultivated, and they are only moderately selected for desirable traits, such as fleshy cladodes or lower spine density (Reyes-Agüero et al. 2005a). For instance, the species O. atropes is endemic to Mexico and is at an incipient stage of domestication, as evidenced by differences in vegetative traits between natural and managed nopaleras (López-Gutiérrez et al. 2015). In the tropical dry forest of Morelos state in central Mexico, O. atropes is used as a food resource by humans, as the young cladodes are highly appreciated (Maldonado-Almanza 2013). In this area, O. atropes is tolerated within agroforestry systems, and nopaleras are frequently found at the intersection of native tropical dry forests and lands that have been transformed for agriculture and livestock uses (Arias-Medellín et al. 2014).

Like many clonally propagated plants, prickly pears are perennial, and hence, they are expected to have preferential outcrossing or even completely self-incompatible mating systems (Barrett et al. 1996; McKey et al. 2010). Because of the disadvantages of clonality (e.g., high inbreeding depression and mate limitation), many clonal domesticated plants still retain sexual recruitment (McKey et al. 2010). Furthermore, repeated outcrossing with their wild counterparts (a frequent phenomenon in Opuntia) precludes the fixation of sterility genes (McKey et al. 2010). Although outcrossing mating systems are expected in Opuntia, there have been documented several Opuntia species with mixed mating systems (Mandujano et al. 2010), as well as species with variation in their mating systems across populations (Ortiz-Martínez et al. 2022). This diversity of mating systems in prickly pears could have evolved in response to environmental pressures such as pollen limitation, resource limitation, and/or the frequency of clonal propagation. High rates of clonal propagation could reduce effective mate availability because many plants are the same genetic individual, leaving few plants with which to outcross; this makes strict outcrossing a less advantageous strategy compared to mixed systems in which selfing also leads to some degree of successful reproduction (Mandujano et al. 2010). In other words, in highly clonal plants, mixed mating systems could be advantageous because geitonogamy (selfing) ensures some reproductive success, while outcrossing promotes recombination (Mandujano et al. 2010). Phenomena such as pollen limitation may also influence the evolution of plants’ sexual expression by favoring the evolution of floral traits that allow self-pollination (Ashman et al. 2004). Pollen limitation has several causes, the most common of which includes invasive plants that interfere with pollination networks, habitat fragmentation, low plant population sizes, and loss of native pollinators (Ashman et al. 2004). These factors are common in disturbed habitats, and hence may ultimately affect plant reproductive systems in those habitats.

Pollination ecology includes such crucial topics as the interactions of the flowers with their visitors and the degree to which plants depend on their pollinators for successful reproduction. Opuntia flowers are diurnal, short-lived, brilliant in color, radially symmetric, and bowl-shaped (Reyes-Agüero et al. 2006; Mandujano et al. 2010). These flowers are mainly bee-pollinated, and their rewards are nectar and pollen. Opuntia flowers are visited by a variety of bee species (mean ± s.d. = 9.4 ± 10.2; data from Tenorio-Escandón et al. 2022), but bee species differ in their effectiveness as pollinators (De Jesús-Romero 2021), and some of them may even act as nectar or pollen robbers (Cota-Sánchez et al. 2013).

Although there have been previous studies of the reproductive biology of Opuntia (Reyes-Agüero et al. 2006; Mandujano et al. 2010), detailed studies of flower-bee interactions would contribute a better understanding of the environmental pressures that may shape successful sexual reproduction and the degree of specialization of these plant species to their pollinators (Tenorio-Escandón et al. 2022). Furthermore, the reproductive system of Opuntia atropes is particularly interesting given that it is a used species subject to management. Specifically, in this study, we sought to determine the mating system of this species and whether their flowers are generalists or specialists in terms of pollination. These data permitted us to discuss the pollination ecology of O. atropes as a native, managed plant in an agroforestry system and reflect on its role in conserving native bee diversity.

Materials and methods

Study sites

We conducted this study in three localities of Morelos state within the municipality of Tlaquiltenango: Valle de Vazquez (18°31'44.66''N, 99°3'51.05''W), Quilamula (18°29'38.94''N, 99° 0'19.38''W), and San Jose de Pala (18°33'3.71'' N, 99° 0'57.23'' W) (^{Fig. S1}). Tlaquiltenango has a warm, subhumid climate with summer rains and a mean annual temperature of 22-24 °C. Annual precipitation varies between 800 and 1,000 mm. The dominant vegetation in Tlaquiltenango is tropical dry forest (46.7%), followed by agricultural lands (31.5%), oak forest (11.3%), pasturelands (9.6%), and urban zones (0.8%) (INEGI 2010). Livestock grazing and agriculture are the predominant economic activities in the area, which have resulted in some degree of disturbance in all three of these localities. The main locality of the study was the village of Valle de Vázquez, where we conducted most of the measurements in a nopalera of O. atropes located along the border between the village and the surrounding tropical dry forest in 2019. The other two localities, Quilamula and San José de Pala, were included to conduct pollination experiments during 2020. In these two latter localities, we worked in nopaleras along the roadsides. In all three nopaleras, we worked with plants of moderate height (between 1-2 m) on which flowers could be manipulated.

Study species

Opuntia atropes Rose is a Mexican endemic species distributed in the West-Central states of Morelos, Mexico, Michoacán, and Guerrero (Britton & Rose 1908; Bravo-Hollis 1978). It inhabits mainly tropical dry forests, pine-oak forests, and xerophytic scrubland (Bravo-Hollis 1978). It has a highly ramified scrub-like life form that reaches up to 3 m in height. Its cladodes are 17-30 cm long and 8-11 cm wide. The cladodes are light green with a pubescent epidermis. The young cladodes collected for human consumption are usually brilliant green with reddish tips. The flowers reach 5 cm in diameter during anthesis. They are yellow, with reddish parts in the most external perianth segments. The pericarpel-the structure that embodies the ovary and later becomes the fruit-is 2 cm wide and 2 cm long. The fruits are globular in shape and green, varying in diameter from 1.5 to 3 cm. Flowering has been recorded in March, and fruiting in April (Bravo-Hollis 1978). Opuntia atropes is appreciated for the quality of its nopalitos (young cladodes used for human consumption) (Reyes-Agüero et al. 2005a; López-Gutiérrez et al. 2015). It is widely used as fodder and living fences (Maldonado-Almanza 2013; Arias-Medellín et al. 2014; Pérez-Sánchez et al. 2015). Like many Opuntia species, O. atropes is clonally propagated, mainly when used for living fences (personal observation). Given the management derived from these uses, O. atropes is at the early stages of an ongoing process of domestication, in which cultivated plants have been found to have larger cladodes, more areoles, and fewer spines than their wild counterparts (López-Gutiérrez et al. 2015).

Reproductive phenology and floral longevity

The flowering and fruiting periods were characterized by visiting Valle de Vázquez every two weeks from February to August 2019. During these visits, we recorded the presence of flower buds, open flowers, and fruits on 50 individuals of O. atropes.

To document individual floral longevity and sexual functions, we selected 20 flowers of O. atropes, each from a different plant. Every two hours, from 10:00 a.m. to 4:00 p.m., we measured flower opening and evaluated the functionality of stigmas and anthers (following Martínez-Peralta et al. 2014b). The flower opening was measured as the diameter of the perianth. The functionality of the stigmas was evaluated as the opening of the stigma lobes (diameter) (Matías-Palafox et al. 2017). Anthers were qualified as functional or dehiscent when their surface was grainy, indicating pollen presentation (Martínez-Peralta et al. 2014b). Dichogamy-the temporal separation of female and male functions-was determined if the stigma and anthers were functional at different intervals of the floral life.

Floral morphology and P/O ratio

We collected 20 flowers in pre-anthesis, each from a different individual, and fixed the flowers in 96% ethanol. Fixed flowers were longitudinally dissected and manipulated under a stereomicroscope to measure the floral traits (Fig. 1B) (Martínez-Peralta et al. 2014b). Before fixation in ethanol, we separated one indehiscent anther per flower and placed it in a 1.5 ml vial. To each vial, we added 1 mL of water and vortexed the vial to facilitate the liberation of the pollen grains from the anther into the water. Immediately after agitation, we used a stereomicroscope to count the pollen grains contained in a 100 µL aliquot of the liquid. The number of pollen grains per aliquot was multiplied by 10 (dilution factor), and this quantity was then multiplied by the total number of anthers of each flower to estimate the number of pollen grains per flower (Martínez-Peralta et al. 2022). We then used this estimated number of pollen grains and the number of ovules per flower to calculate the pollen/ovule ratio (P/O) per flower.

Figure 1
Flowers of Opuntia atropes in a tropical deciduous forest of central Mexico. A. Open flower and flower bud on the cladode. B. Floral measurements considered for floral morphology. C. The solitary bee Diadasia sp. after landing on the flower's stigma, and then (D) enters the male whorl. Floral measurements in panel B: (A) Perianth width, (B) floral length, (C) Number of perianth segments, (D) Number of stamens, (E) Stamen length, (F) Polar diameter of ovary chamber, (G) Equatorial diameter of ovary chamber, (H) Number of stigma lobes, (I) Style-stigma length, (J) Number of ovules.

We used the P/O ratio approach to infer the breeding system (Cruden 1977). Plant species with the highest P/O ratios are expected to have an obligate xenogamous mating system (i.e., require strict outcrossing to set fruits). In contrast, those with the lowest P/O ratios have cleistogamous flowers (i.e., flowers that do not open at all and do not require pollen transfer for setting fruits) (Cruden 1977). The following cutoff values were used to define breeding system categories: P/O of 5859 ± 936 = xenogamy, 796 ± 87 = facultative xenogamy, 169 ± 22 = facultative autogamy, 28 ± 3 = obligate autogamy, and 5 ± 0.1 = cleistogamy. We also calculated an outcrossing index (OCI) based on three floral traits: perianth diameter, herkogamy and dichogamy. OCI values ranges from 0 to 4 points, assigned as follows: 1 point for the presence of herkogamy, 1 point for the presence of dichogamy, and between 0 and 3 points according to perianth size (perianth size < 1mm = 0, 1-2 mm wide = 1, 2-6 mm = 2, and wider than 6 mm = 3) (Cruden 1977). Based on this punctuation, breeding system is classified as follows: cleistogamy (0), autogamy (1), facultative autogamy (2), facultative xenogamy (3), and obligate xenogamy (4) (Cruden 1977).

Mating system and pollen limitation

To determine the mating system of O. atropes, we conducted five pollination treatments (Kearns & Inouye 1993; Martínez-Peralta & Mandujano 2011) in Valle de Vazquez through March 2019. For the Autonomous Selfing (ASELF) treatment, flowers were covered and otherwise unmanipulated throughout the anthesis to test for autonomous intraflower self-pollination. In the Manual Selfing (SELF) treatment, flowers were covered during the pre-anthesis and then manually pollinated with pollen from the same flower. The Outcrossing (CROSS) treatment consisted of covering the flowers during pre-anthesis to prevent insect visitation and manually cross-pollinating them using a mixture of pollen from five donor plants. In the Natural Pollination (NAT) treatment, flowers were left uncovered and available to pollinators throughout the anthesis. Finally, in the Supplementary Pollen (SUPL) treatment, flowers were left available to pollinators and were also manually pollinated once with a mixture of pollen from five donor plants. Each treatment comprised 20 flowers, each from a different individual. For the treatments SELF, CROSS, and SUPL, manual pollination was conducted between noon and 14:00, the time of maximum stigma receptivity. Pollen donors for the CROSS and SUPL treatments were plants located in the surroundings of the nopalera, away from the recipient plants, to minimize the risks of inbreeding. At the end of the anthesis, we covered each flower with a mesh bag that was tied to the cladode to prevent any harm to the developing fruits from cattle, herbivores, or humans. During July 2019 many plants in the nopalera were cut down by local people before some of the fruits were collected, and the SUPL treatment was completely lost. To overcome this situation, we repeated the NAT and SUPL treatments in 2020, using plants from San José de Pala and Quilamula. These two pollination treatments were conducted in the same year to be able to test for pollen limitation.

Fruits were considered to be mature when their coloration was reddish; they were collected between May and July 2019 and during July 2020. For each fruit we recorded weight (g), length (mm), width (mm), and number of seeds. We evaluated fruit set across treatments using a generalized linear model (GLM) with binomial error and logit link function and seed set using a GLM with Poisson error and log link function (Dunn & Smyth 2018). After conducting these two GLMs, we conducted orthogonal contrasts to determine pairwise differences among treatments (Mangiafico 2015). Fruit weight, length, and width were analyzed using ANOVA, then post-hoc Tukey comparisons were conducted to determine differences among treatments. All analyses were conducted in the program R (R Core Team 2020).

As an indicator of the mating system, we obtained the outcrossing index t _e = w _s/(w _x+w _s), where w _x is the progeny from outcrossing and w _s is the progeny from selfing (Mandujano et al. 2010). For this index, progeny means the proportion of fruits (fruits/flowers) and proportion of seeds (seeds/ovules) obtained from pollination treatments CROSS and SELF (for w _x and w _s, respectively). This index ranges from 0 to 1, expressing the relative contribution of the progeny from outcrossing in the total plant fitness. An outcrossing index of 0 refers to strict selfing, 1 to strict outcrossing, and values between 0.2 and 0.8 represent mixed mating systems (Mandujano et al. 2010).

To determine the extent of pollen limitation, we used the index of pollen limitation, L, proposed by Larson and Barrett (2000): L = 1 - (P _o / P _s), where P _o is the relative success of natural pollination, and P _s the relative success of supplemental pollination. This index was calculated using the fruit and seed set from the pollination treatments conducted during 2020.

Pollinators

To determine the identities of floral visitors of O. atropes, we conducted direct observations of 20 flowers from 5 plants (4 flowers per plant). These 20 flowers were observed for 45 min, followed by a 15 min break for each hour between 10:00 a.m. and 3:00 p.m., resulting in a total sampling time of 270 min (4.5 h) per flower. During these observations, we considered a floral visitor as any insect or vertebrate that made physical contact with the flower. We recorded the taxonomic identity of the floral visitor, the part of the flower it had touched (stamens, stigma, both organs, or the perianth segments), the handling time (the time that a visitor spends in a flower) (Ne’eman et al. 2006), and its behavior on the flower. For each species of visitor, we calculated the contact rate as the percentage of visits in which it contacted stamens, stigma, or both organs, considering that pollinators contact both organs in a single visit (Sigrist et al. 2021). One specimen of each species of floral visitor was photographed under a stereomicroscope; these images were analyzed with ImageJ (Rasband 2018) to determine the body length of floral visitors to the nearest mm. Body size was used to discuss the effectiveness of each floral visitor as a pollinator. To identify the insect floral visitors, we collected some individuals directly from flowers by sacrificing them in ethyl acetate killing jars, and then pinning the specimen. Bees were identified by comparing with bees that were previously collected and identified from the same study sites (Martínez-Peralta et al., in preparation) and from the Colección Entomológica del Centro de Estudios en Zoología (CZUG) at the University of Guadalajara. We calculated the visitation rate of each species of visitor as the total visits/total observation time. Species with less than 10 total visits were pooled into the category “Other”. To determine whether the species differed in handling time, we conducted a non-parametric Kruskal-Wallis test (Mangiafico 2015); pairwise differences between species were identified using a post-hoc test with Holm-Bonferroni correction for multiple comparisons (Mangiafico 2015).

Results

Reproductive phenology and floral longevity

We found O. atropes buds beginning in early February and flowers in anthesis from the last week of February to the last week of May. During May, the number of open flowers decreased markedly, while some floral buds remained unopen. Immature fruits were found beginning in May, and the fruiting period lasted from May to July.

Opuntia atropes had hermaphroditic, actinomorphic flowers (Fig. 1) that were similar to other Opuntia flowers in morphology and size. They were bowl-shaped, with a perianth width of 29.7 ± 10.5 mm (mean ± s.d.) and a floral length of 44.2 ± 16.0 mm (this length includes perianth and pericarpel length). The pistils protruded from the center of the male whorl (Fig. 1A, B) and measured 20.2 ± 7.0 mm in length. The ovular chamber measures 5 ± 2.2 mm long and 3.9 ± 1.5 mm wide, containing 152 ± 133 ovules. There were 430 ± 195 stamens per flower, measuring 14.8 ± 5.2 mm in length. The number of pollen grains per flower was 177 760 ± 90 766.

Opuntia atropes flowers were diurnal, with a longevity of 6 h. They opened from 10:00 a.m. to 4:00 p.m., with a maximal opening at 12:00 p.m. Stigma opening, a proxy of stigma receptivity, reached the widest diameter at 2:00 p.m. The anthers began to shed pollen before the stigma became receptive, around 10:00 a.m.; however, both functions occurred simultaneously around 12:00 p.m., so these flowers are homogamous. There was a significant difference between the mean length of stamens versus stigmas (t = 11.45, d.f. = 19, P = 5.73e-10); therefore, flowers of O. atropes were herkogamous.

Flowers of O. atropes showed changes in color throughout their life. At the anthesis beginning, flowers were categorized within a single category of the Munsell system (5Y/8/10) (^{Fig. S2}). At 12:00 p.m., aside from the 5Y hue, the hue 2.5Y was also recorded. Flowers in color categories that included hues between yellow and red were recorded from 2:00 p.m. to 4:00 p.m. In summary, these results indicate that floral color changed from a highly uniform yellow hue at the beginning of anthesis to more heterogeneous categories of hue, chroma, and value.

The flowers of O. atropes were wider than 3mm and thus received a flower size score of 3 in the OCI (Cruden 1977); they are herkogamous, which was scored as 1 point, and homogamous (with sexual functions occurring simultaneously), scored as 0 points. Thus, the overall OCI score was 4, corresponding to obligate xenogamy (Cruden 1977). The pollen/ovule ratio (P/O) obtained was 1,418 ± 873 (mean ± standard deviation), which corresponds to a facultative xenogamous system according to the categorization proposed by Cruden (1977).

Mating system and pollen limitation

There were significant differences among pollination treatments in fruit and seed production (Z = -2.87, d.f. = 139, P = 0.0041 and Z = 21.034, d.f. = 76, P = 2e-16, respectively). All treatments that involved natural and cross-pollination set significantly more fruits than the two selfing treatments (Fig. 2). Fruit weight, length and width also differed significantly among treatments (F =18.71, d.f. = 5, P = 4.49e-14; F = 34.14, d.f. = 5, P = 2e-6; and F = 30.8, d.f. = 5, P = 2e-016, respectively). Due to collinearity in these three fruit measurements, only fruit weight is shown. Natural pollination from 2019 (NAT) and supplemental pollination had the heaviest fruits, followed by natural pollination from 2020 (NAT2) and the cross-pollination treatment; finally, both the manual and autonomous selfing treatments set the lightest fruits (Fig. 2). The outcrossing rate obtained indicates a prevalence for outcrossing in fruit set (t _e = 0.95), but outcrossing is not as strict for seed set (t _e = 0.66); however, the number of seeds to obtain the t _e comes from a unique fruit. These results indicate that in O. atropes, outcrossing is more effective than selfing.

Figure 2
Fruit set (A), seed set (B), and fruit weight (C) following the pollination treatments conducted in O. atropes.

Natural pollination and supplementary pollen conducted in 2020 (NAT2 and SUPL2, respectively) were statistically indistinguishable in the percent fruits set. However, the supplementary pollination treatment (SUPL2) set significantly more seeds than the natural pollination treatment (NAT2), suggesting pollen limitation for the seed set. The pollen limitation index value suggested a moderate pollen limitation for fruits (L = 0.15) and seeds (L = 0.17).

Pollinators

The observed floral visitors of O. atropes comprised nine species of bees, one wasp, and one hummingbird species. The solitary bee Diadasia sp. accounted for most visits, with 74.2% of relative frequency (RF). Other visiting bees were Augochlorella sp. (6.3%), Lithurgopsis (6.3%), one morphospecies of Halictidae (that was not collected and referred to as Halictidae) (4%), Apis mellifera (2.6%), Xylocopa sp. (2.6%), Ashmeadiella sp. (1.1%), Euglossa sp. (not collected) (0.6%), and Ceratina sp. (0.15%). The wasp (Vespidae) and the hummingbird (Trochilidae) represented 0.5% and 1.7% of the visits, respectively. Flowers were visited between 10:00 am and 4:00 pm, with a peak activity between 12:00 p.m. and 1:00 p.m. (26% of the total visits). The number of visits per flower during the observation period ranged from 15-67, with a mean visitation rate of 3 ± 7.6 per hour (mean ± s.d.). During our observations of floral visitors, we detected that stamens have thigmonastic movements: they move towards the style when touched by floral visitors.

During most of the visits, floral visitors made contact with at least one sexual organ. The overall contact rate was 44.5% for stamens, 16.2% for stigmas, and 35.8% for both organs. Of the total visitors recorded, only 1.8% touched only the perianth segments. All of the visits conducted by the hummingbird (1.7%) were focused on searching for nectar, and we did not observe clear contact with the flower’s sexual organs. When visiting a flower, the solitary bee Diadasia sp. most frequently landed on the stigma and then descended to the male whorl, often reaching the inner stamens (Fig. 1). This species had the highest contact rate of both organs during a single visit (Table 1). Lithurgopsis sp. landed on the stigma in some visits, but this was not as consistent as in Diadasia sp., and the rate of contact with the stamens was higher than with the stigma or both organs. In contrast, the most frequent contact of Ashmeadiella sp., Halictidae, and Xylocopa sp. was with the stamens, suggesting they primarily collected pollen. Despite being able to contact both organs at a similar rate as Diadasia sp., Apis mellifera did not show the characteristic pattern of landing on the stigma shown by Diadasia sp. Due to the remaining visitors' low frequency, it was impossible to describe clear behavior patterns.

Table 1
Rate of visitation, contact rate, handling time, and body size of floral visitors of O. atropes in March 2019 (n = 20 flowers) in Tlaquiltenango, Morelos, a locality of tropical dry forest in central Mexico. Different letters in Handling time column indicate significant differences among groups following the pairwise comparisons after the Kruskal-Wallis test. Species with an asterisk in Handling time column were grouped as the category “Other” due to low frequency of visit.

Floral visitor	Rate of visitation (mean ± SD) (visits/hour)	Stamens contact rate (%)	Stigma contact rate (%)	Stamens and stigma contact rate (%)	Handling time (s)	Body size (length) (mm)
Diadasia sp.	5.36 ± 2.31	37.3	19.9	41.1	13.3 ± 19.3 a	12
Augochlorella sp.	0.48 ± 0.74	51.2	7	39.5	15.9 ± 24 ab	6
Lithurgopsis sp.	0.44 ± 0.11	82.5	2.5	12.5	9.75 ± 12.3 a	11
Halictidae	0.29 ± 0.10	84.6	0	15.4	14.1 ± 23.1 a	NA
Apis mellifera	0.20 ± 0.05	55.6	5.6	38.9	7.5 ± 6.4 ab	11
Xylocopa sp.	0.17 ± 0.09	80	13.3	6.7	6.13 ± 5.4 bc	21
Trochilidae	0.12 ± 0.05	0	0	0	3.55 ± 2.5 bc	NA
Ashmeadiella sp.	0.08 ± 0.04	85.7	0	0	6.1 ± 6.2* bc	7
Euglossa sp.	0.04 ± 0.03	75	25	0	3.25 ± 1.9* bc	NA
Vespidae	0.03 ± 0.02	33.3	33.3	33.3	4.67 ± 0.6* bc	9
Ceratina sp.	0.01 ± 0.01	0	0	0	2* bc	7

The handling time of all visitors together ranged from 1 to 120 s with a positive-skewed distribution. The non-parametric Kruskal-Wallis test indicated significant differences in handling times among groups (chi-square = 18.8, d.f. = 7, P = 0.0088). According to the pairwise comparison with multiple comparison correction, Diadasia sp., Halictidae, and Lithurgopsis sp. had similar handling times to A. mellifera and Augochlorella sp. but spent significantly more time in a single visit than Trochilidae, Xylocopa sp., and the low-frequency floral visitors grouped as “Other” (Table 1). These data of floral visitors indicate that Diadasia sp. is the species that more effectively pollinates the flowers of O. atropes based on frequency and behavior.

Discussion

Flowering and fruiting in O. atropes occur simultaneously, a phenological reproductive pattern described in other cacti as asynchronous (Pimienta-Barrios & Nobel 1998; Bustamante 2003; Razo-León et al. 2022). Flowering of O. atropes occurs simultaneously with O. decumbens, another species of prickly pear that occurs in the same area (Gómez-Aguilar 2022). Flowering of O. atropes occurs in the middle of the dry season of the tropical dry forest, when floral resources from native vegetation are limited (Hernández-Galindo 2016). It thus may represent an essential source of floral resources (pollen and nectar) for native bees. Fruiting also occurs during the dry season and, hence, could constitute an important food resource for vertebrate fauna, such as rodents and birds (González-Espinosa & Quintana-Ascencio 1986).

Floral longevity of O. atropes is relatively short (6-7 h), a feature that can be considered advantageous since it limits the loss of water from transpiration in a dry environment (Primack 1985; Pimienta-Barrios 1990). Color variation throughout the floral lifespan could be linked to either the age of the flower per se or to the process of pollination (Erickson & Pessoa 2022), which have both been documented in other Opuntia (Pimienta-Barrios 1990; Piña et al. 2007). Our observations indicate that pollination would be a plausible mechanism underlying floral color variation since bagged flowers of the autonomous selfing treatment remained closer to yellow hues (personal observation). In the genus Opuntia, whose fruits are edible and constitute an important economic resource, it would be helpful to determine whether flower color changes after pollination, therefore, could be used as an easy proxy of successful pollination.

We found that in O. atropes, sexual reproduction is an important mode of propagation in the agroforestry system where the populations were located. Sexuality can be lost in clonally propagated plants due to a negative feedback loop that diminishes the frequency of fertile clones (McKey et al. 2010). However, such a loss of sexuality in Opuntia has been rarely documented (Negrón-Ortiz 1998). According to the pollination experiment, O. atropes has a predominantly outcrossing mating system and may express some degree of self-incompatibility, given the low success of the selfing pollination treatments (ASELF and SELF). These results indicate that this species depends on native bees to set fruits. Across Opuntia, species with mixed mating systems are more common than those with strict or predominantly outcrossing systems (Mandujano et al. 2010). Floral attributes that promote this outcrossing mating system are floral herkogamy, large flowers, and the high pollen/ovule ratio. Although flowers are homogamous, herkogamy could be sufficient in promoting outcrossing and avoiding self-interference (Barrett & Harder 2006). The fact that there was a minimal but non-zero success of selfing indicates a slight departure from strict outcrossing in O. atropes and could indicate mating system evolution (Vogler & Kalisz 2001; Goodwille et al. 2005). For example, a population of O. macrocentra that occurs in an environment with pollinator availability was found to be completely self-incompatible. In contrast, a different population in a site with poor pollinator availability had a mixed mating system (Ortiz-Martínez et al. 2022). Partial self-incompatibility, in which some individuals are self-compatible while others are self-incompatible in the same population (Levin 1996; Ferrer et al. 2009), has been reported in the cactus family and was linked to pollen limitation (Martínez-Peralta et al. 2014a; Martínez-Peralta & Mandujano 2016). These two cases of cacti that slightly deviate from strict outcrossing may represent a transition to mixed mating systems. Mixed mating systems are thought to evolve as a strategy for reproductive assurance (Goodwillie et al. 2005) in poor pollinator environments or when compatible mates are absent. In the studied populations of O. atropes, the high levels of human disturbance from livestock and agriculture may lead to unpredictable pollinator availability. Outcrossing in a species with clonal propagation and a large floral display could face processes that decrease effective pollination, such as intraplant or intraflower pollination due to recurrent visits (i.e., geitonogamy). The high success rate of natural pollination in this study suggests that native bees, in addition to having good within-flower performance, also transport pollen effectively among flowers due to their foraging pattern (Schlindwein 2004).

Although there were no statistically significant differences in fruit sets among treatments that involved outcrossing (Natural Pollination (NAT, NAT2), Outcrossing (CROSS), and Supplementary pollen (SUPL)), fruit sets tended to be lower in both treatments conducted during the second year (NAT2 and SUPL). However, there were differences in seed set and fruit weight between Natural pollination treatments conducted in different years (NAT and NAT2). Despite the mixed effect of year and site, these comparisons suggest variability in reproductive success in time and/or space. The variability of reproductive success is multifactorial and could involve maternal effects, availability of pollinators, stigmatic pollen load, and climatic variation, among other factors (Burd et al. 2009), and it may affect floral traits. Among the floral traits we measured, there was notably wide variation in the number of ovules per flower (coefficient of variation = 82.4). Ovule number variation has been reported to be related to an unpredictable pollination environment; when ovule fertilization is highly variable among flowers within a population, the high variation in ovule number permits some flowers to be very successful, even when though some resources were wasted on the numerous non-pollinated flowers (Burd et al. 2009). The high variability in fertilization could also be related to stigmatic pollen loads, which reflect the efficacy of pollen transfer in the population. A better understanding of the relationship between ovule number variation and stigmatic pollen loads would be informative of the extent of reproductive success in the population and how reproductive success affects the evolution of floral traits in O. atropes.

We found moderate pollen limitation in O. atropes, as seed set and fruit weight were higher for the supplemental pollination treatments (conducted in 2020). Our observations of floral visitors conducted in 2019 and the success of all naturally pollinated flowers during that year suggest that pollinators had an optimal performance for fruit and seed formation in that particular year and site. However, the supplementary pollen treatment, conducted in a different year in another locality, indicated that conditions for successful pollination are likely to change and that pollen limitation may arise in O. atropes. Pollen limitation has been documented in many angiosperms (Ashman et al. 2004), including cacti (Piña et al. 2007; Martínez-Peralta & Mandujano 2016). Pollen limitation is linked to low mate availability, with two leading proximate causes: a low number of plants to outcross and low pollinator abundance or performance (Ashman et al. 2004). In this species, low mate availability could result from the typical clonal propagation of Opuntia. Clonal propagation occurs naturally through cladode fragmentation and human management, as this species is widely used. Suboptimal pollinator performance cannot be ruled out since pollinator assemblages fluctuate in time and space (Herrera 1988). Based on these results, we recommend a more detailed analysis of the extent of pollen limitation in O. atropes, including temporal and spatial variability in reproductive success, to disentangle which factors account for pollen limitation.

Floral visitors to O. atropes were mainly bees, although a hummingbird and a wasp were also observed. This diversity of visitors indicates that O. atropes has the potential to attract a broad assemblage of floral visitors from different taxonomic groups. These observations and the floral morphology could suggest an apparent and phenotypic generalization of the flower based on floral traits (sensuOllerton et al. 2007). In other words, the apparent generalized floral morphology (large, radially symmetrical flowers) would permit a high diversity of floral visitors to interact with the flower. However, our findings also suggest that O. atropes flowers may be more specialized than they appear since most pollination (both in terms of rate of visitation and pollinator efficiency) could be attributed to a single bee species, Diadasia sp. This species accounted for the majority of visits (74%) and touched both sexual organs at a higher rate than the other visitors. The high contact rate of both organs could be promoted by its behavior of landing on the stigmas (Fig. 1) and its size (Table 1), as described for other Opuntia species in which effective pollinators are medium to large bees (Fachardo & Sigrist 2020). This solitary bee belongs to a genus whose preferred host is Opuntia (Sipes & Tepedino 2005; Ortiz-Martínez et al. 2022). In O. atropes, the halictids (Augochlorella sp. and the morphospecies of Halictidae) seem inefficient pollinators, since they spend long handling times on the flower, even though they do not reach the inner stamens to collect pollen. The small body size of these halictids could also be linked to the low contact rate of both organs. Similar to the halictids, A. mellifera and Lithurgopsis sp. spent substantial time on the flowers, and despite their size, their contact rates suggest that their behavior does not favor pollination. The fact that Diadasia sp. had similar handling times to other visitors that were less frequent and with low rates of contact with both sexual organs, indicates that handling time does not necessarily correlate to performance as a pollinator. In addition, thigmonastic stamens could prevent inefficient pollinators from accessing the nectar chamber and the innermost and most rewarding anthers (Cota-Sánchez et al. 2013). These results indicate that O. atropes is specialized for bee pollination, conducted by an Opuntia specialist bee, as has been documented for other species of prickly pears (Ortiz-Martínez et al. 2022). Floral traits, like the thigmonastic stamens, provide further support for the idea of phenotypic specialization of the flower (Ollerton et al. 2007) as documented in the epiphytic cactus Rhipsalis neves-armondii (Martins & Freitas 2018).

Although the other visiting bees were less effective pollinators than Diadasia sp., their interactions with O. atropes flowers would provide them with floral resources such as pollen and nectar. For instance, as the contact rates revealed, all bees except Diadasia sp. and A. mellifera touched only the stamens on most of their visits, which suggests that these bees visit O. atropes mainly to collect pollen. Our observations indicate that the hummingbird would visit O. atropes flowers for nectar consumption, rather than pollinating. The flowering period of O. atropes occurs during the dry season when other floral resources are scarce, and bee abundance is low (Martínez-Peralta et al., in preparation); this scenario makes O. atropes an important floral resource for native bees and possibly other nectar consumers, because of their abundant pollen, nectar, large floral display, and extended flowering period. When abundant floral resources appear, or their preferred hosts are scarce, bees may abandon their usual foraging patterns to take advantage of the available resources (Waser 1986).

It is noteworthy that even though the main study site was located at the edge of the tropical dry forest and a village and is subject to chronic disturbance from agriculture and cattle grazing, the most frequent visitor to O. atropes flowers was a native Opuntia specialist bee, Diadasia sp. (Sipes & Tepedino 2005). Overall, the exotic bee Apis mellifera accounted for a meager percentage of the visits (2.6%), which indicates that the pollination system occurs mainly between native bees and O. atropes, a native plant. Other studies in the area have shown that although present due to apiculture, A. mellifera is not a frequent visitor of cactus flowers (Martínez-Peralta & Martínez-Zavala 2021; Gómez-Aguilar 2022). Thus, although cactus species are not dominant in the tropical dry forest, they constitute valuable floral resources for native bees. The main rewards that Opuntia species offer to visiting bees are nectar and pollen. Nectar is often produced copiously (Reyes-Agüero et al. 2006; Mandujano et al. 2010) and actively consumed by adult bees. Pollen is abundant and highly nutritious (Osborn et al. 1988); it is collected by female bees to provide food to their larvae (Michener 2007). Therefore, aside from the management of O. atropes for human consumption, this plant is ecologically relevant since it is visited by an important number of native bee species and offers them pollen, nectar, and even refuge. Other studies have documented that bee diversity is high in vegetation where Opuntia is dominant (Sánchez-Echeverría et al. 2016; Ávila-Gómez et al. 2019), suggesting that nopaleras may act as reservoirs of wild bee diversity.

We found that O. atropes has an outcrossing mating system, which is promoted by floral size and herkogamy, and may have a self-incompatibility system (Goodwillie et al. 2005). Despite its clonal growth and human management, sexuality in this species is fundamental to the maintenance of populations. Our data on seed number and fruit weight indicate that reproductive output is pollen-limited, and the extent of this pollen limitation may be variable in time and/or space. The most effective pollinator is the native bee Diadasia sp., which, together with the floral morphology and the thigmonastic stamens, indicates a functionally specialized pollination. This study highlights the importance of studying the pollination ecology of Opuntia species. In addition to their cultural and economic significance, Opuntia species have a critical role as a source of floral resources for native bees, which further increases their ecological importance in the context of pollinator declines.

Acknowledgments

We thank V. Cazares-Soriano, A. Palma-Bahena, L. Hernández-Bautista, and H. Armenta-Gómez for their valuable help during fieldwork. We thank A. G. Trejo-Loyo, E. Cuevas-García, and R. E. Alcalá-Martínez for previous manuscript revisions. We thank Lynna Kiere for English revision.

References

• Arias-Medellín LA, Flores-Palacios A, Martínez-Garza C. 2014. Cacti community structure in a tropical Mexican dry forest under chronic disturbance. Botanical Sciences 92: 405-415.
• Ashman T, Knight TM, Steets JA et al 2004. Pollen limitation of plant reproduction: Ecological and evolutionary causes and consequences. Ecology 85: 2408-2421.
• Ávila-Gómez ES, Meléndez-Ramírez V, Castellanos I, Zuria I, Moreno CE. 2019. Prickly pear crops as bee diversity reservoirs and the role of bees in Opuntia fruit production. Agriculture, Ecosystems & Environment 279: 80-88.
• Barrett SCH, Harder LD, Worley AC. 1996. The comparative biology of pollination and mating in flowering plants. Philosophical Transactions of the Royal Society London B 351: 1271-1280.
• Barrett SCH, Harder LD. 2006. David G. Lloyd and the evolution of floral biology: from natural history to strategic analysis. In: Harder LD, Barrett SCH (eds.). Ecology and Evolution of Flowers. New York, Oxford University Press. p. 1-21.
• Britton NL, Rose N. 1908. A preliminary treatment of the Opuntioideae of North America. Smithsonian Miscellaneous Collections 50: 503-539.
• Burd M, Ashman TL, Campbell DR et al 2009. Ovule number variation per flower in a world of unpredictable pollination. American Journal of Botany 96: 1159-1167.
• Bustamante E. 2003. Variación especial y temporal en la reproducción y estructura poblacional de Stenocereus thurberi: Una cactácea columnar del matorral costero del sur de Sonora, México. MSc Thesis, Universidad Nacional Autónoma de México, Mexico City.
• Bravo-Hollis H. 1978. Las Cactáceas de México. Mexico City, Universidad Nacional Autónoma de México. vol. I.
• Carman K, Jenkins D. 2016. Comparing diversity to flower-bee interaction networks reveals unsuccessful foraging of native bees in disturbed habitats. Biological Conservation 202: 110-118.
• Cota-Sánchez JH, Almeida OJG, Falconer DJ, Choi HJ, Bevan L. 2013. Intriguing thigmonastic (sensitive) stamens in the Plains Prickly Pear Opuntia polyacantha (Cactaceae). Flora 208: 381-389.
• Cruden RW. 1977. Pollen-ovule ratios: A conservative indicator of breeding systems in flowering plants. Evolution 31: 32-46.
• De Jesús-Romero Y. 2021. Eficiencia de los polinizadores de Opuntia robusta Wedland (Cactaceae) en el Jardín Botánico Regional de Cadereyta, Querétaro, México. Bachelor’s Thesis, Universidad Nacional Autónoma de México, Mexico City.
• Dunn PK, Smyth GK. 2018. Generalized Linear Models with examples in R. New York, Springer.
• Erickson MF, Pessoa DM. 2022. Determining factors of flower coloration. Acta Botanica Brasilica 36: e2021abb0299.
• Fachardo ALS, Sigrist MR. 2020. Pre-zygotic reproductive isolation between two synchronopatric Opuntia (Cactaceae) species in the Brazilian Chaco. Plant Biology 22: 487-493.
• Fernández-Montes MR, Mondragón-Jacobo C, Luna-Vázquez J et al 2000. Principales cultivares mexicanos de nopal tunero. Publicación Técnica Núm. 1. INIFAP-CIRCE-Campo Experimental Norte de Guanajuato, México.
• Ferrer M, Good-Ávila SV, Montaña C, Domínguez D, Eguiarte LE. 2009. Effect of variation in self-incompatibility on pollen limitation and inbreeding depression in Flourensia cernua (Asteraceae) scrubs of contrasting density. Annals of Botany 103: 1077-1089.
• Gómez-Aguilar KS. 2022. Ecología reproductiva de Opuntia decumbens Salm-Dyck en la Reserva de la Biosfera Sierra de Huautla, Morelos, México. Bachelor’s degree Thesis, Universidad Nacional Autónoma de México Mexico City.
• Gómez-Otamendi E, Ortiz-Arteaga Y, Ávila-Gómez ES, Pérez-Toledo G, Valenzuela J, Moreno CE. 2018. Diversidad de hormigas epigeas en cultivos de nopal tunero (Opuntia albicarpa) y matorrales de Opuntia spp. del estado de Hidalgo, México. Revista Mexicana de Biodiversidad 89: 454-465.
• González-Espinosa M, Quintana-Ascencio PF. 1986. Seed predation and dispersal in a dominant desert plant: Opuntia, birds, and mammals. In: Estrada A, Fleming TH (eds.). Tasks for vegetation science: Vol. 15 Frugivores and seed dispersal. Dordrecht, Springer. p. 273-284.
• Goodwillie C, Kalisz S, Eckert CG. 2005. The evolutionary enigma of mixed mating systems in plants: occurrence, theoretical explanations, and empirical evidence. Annual Review of Ecology, Evolution, and Systematics 36: 47-79.
• Griffith MP. 2004. The origins of an important cactus crop, Opuntia ficus-indica (Cactaceae): New molecular evidence. American Journal of Botany 91: 1915-1921.
• Hernández-Galindo MA. 2016. Fenología floral, expresión sexual de árboles y visitantes florales en sitios de restauración ecológica experimental en Sierra de Huautla, Morelos. Bachelor’s Thesis, Benemérita Universidad Autónoma de Puebla, Puebla.
• Herrera CM. 1988. Variation in mutualisms: The spatio-temporal mosaic of a pollinator assemblage. Biological Journal of the Linnean Society 35: 95-125.
• INEGI - Instituto Nacional de Estadística y Geografía. 2010. Compendio de información geográfica municipal 2010. Tlaquiltenango, Morelos. https://www.inegi.org.mx/contenidos/app/mexicocifras/datos_geograficos/17/17025.pdf 23 May 2023.
  » https://www.inegi.org.mx/contenidos/app/mexicocifras/datos_geograficos/17/17025.pdf
• Kearns CA, Inouye DW. 1993. Techniques for pollination biologists. Colorado, University Press of Colorado.
• Larson BMH, Barrett SCH. 2000. A comparative analysis of pollen limitation in flowering plants. Biological Journal of the Linnean Society 69: 503-520.
• Levin DA. 1996. The evolutionary significance of pseudo-self-fertility. The American Naturalist 148: 321-332.
• López-Gutiérrez DM, Reyes-Agüero JA, Muñoz A, Robles J, Cuevas E. 2015. Comparación morfológica entre poblaciones silvestres y manejadas de Opuntia atropes (Cactaceae) en Michoacán, México. Revista Mexicana de Biodiversidad 86: 1072-1077.
• Maldonado-Almanza BJ. 2013. Patrones de uso y manejo de los recursos florísticos del bosque tropical caducifolio en la Cuenca del Balsas, México. PhD Thesis, Universidad Nacional Autónoma de México, Mexico City.
• Mandujano MC, Carrillo-Angeles I, Martínez-Peralta C, Golubov J. 2010. Reproductive biology of Cactaceae. In: Ramawat KG (ed.). Desert Plants - Biology and Biotechnology. Berlim, Springer. p. 197-230.
• Mangiafico SS. 2015. An R Companion for the Handbook of Biological Statistics, version 1. 3. 2. https://rcompanion.org/rcompanion/a_02.html 10 Feb. 2023.
  » https://rcompanion.org/rcompanion/a_02.html
• Martínez-Peralta C, Gonzaga-Segura JA, Arroyo-Cosultchi G. 2022. Floral biology of two woody species of Dalbergia at high risk of timber extraction. Brazilian Journal of Botany 45: 743-753.
• Martínez-Peralta C, Márquez-Guzmán JG, Mandujano MC. 2014a. How common is self-incompatibility across species of the herkogamous genus Ariocarpus? American Journal of Botany 101: 530-538.
• Martínez-Peralta C, Molina-Freaner F, Golubov J, Vázquez-Lobo A, Mandujano MC. 2014b. A comparative study of the reproductive traits and floral morphology of a genus of geophytic cacti. International Journal of Plant Sciences 75: 663-680.
• Martínez-Peralta C, Mandujano MC. 2011. Reproductive ecology of the endangered living rock cactus, Ariocarpus fissuratus (Cactaceae). Journal of the Torrey Botanical Society 138: 145-155.
• Martínez-Peralta C, Mandujano MC. 2016. Pollen limitation in a rare cactus with synchronous mass flowering. Haseltonia 22: 2-8.
• Martínez-Peralta C, Martínez-Zavala A. 2021. Flower biology of the cactus Coryphantha elephantidens in the tropical dry forest of central Mexico. Plant Species Biology 36: 102-113.
• Martins C, Freitas L. 2018. Functional specialization and phenotypic generalization in the pollination system of an epiphytic cactus. Acta Botanica Brasilica 32: 359-366.
• Matías-Palafox MA, Jiménez-Sierra CL, Golubov J, Mandujano MC. 2017. Reproductive ecology of the threatened “star cactus” Astrophytum ornatum (Cactaceae): A strategy of continuous reproduction with low success. Botanical Sciences 95: 245-258.
• McKey D, Elias M, Pujol B, Duputié A. 2010. The evolutionary ecology of clonally propagated domesticated plants. New Phytologist 186: 318-332.
• Michener CD. 2007. The bees of the world. Maryland, Johns Hopkins University Press.
• Ne’eman G, Shavit O, Shaltiel L, Shmida A. 2006. Foraging by male and female solitary bees with implications for pollination. Journal of Insect Behavior 19: 383-401.
• Negrón-Ortiz V. 1998. Reproductive biology of a rare cactus, Opuntia spinosissima (Cactaceae), in the Florida Keys: Why is seed set very low? Sexual Plant Reproduction 11: 208-212.
• Ollerton J, Killick A, Lamborn E, Watts S, Whiston M. 2007. Multiple meanings and modes: On the many ways to be a generalist flower. Taxon 56: 717-728.
• Ortiz-Martínez LE, del Castillo RF, Mandujano MC. 2022. Variability in mating strategies in a widespread cactus in the Chihuahuan Desert. Nordic Journal of Botany 12: e03806.
• Osborn MM, Kevan PG, Lane MA. 1988. Pollination biology of Opuntia polyacantha and Opuntia phaeacantha (Cactaceae) in southern Colorado. Plant Systematics and Evolution 159: 85-94.
• Pérez-Sánchez RE, Delgado-Sánchez LA, García-Saucedo PA, Pulido J, Ortiz-Rodríguez R. 2015. Caracterización, modelación morfológica y análisis proximales de Opuntia ficus-indica y O. atropes durante las épocas de estiaje y lluvias. Nova Scientia 7: 133-152.
• Pimienta-Barrios E. 1990. El nopal tunero. Mexico, Departamento de Investigación Científica y Superación Académica de la Universidad de Guadalajara.
• Pimienta-Barrios E, Nobel PS. 1998. Vegetative, reproductive, and physiological adaptations to aridity of pitayo (Stenocereus queretaroensis, Cactaceae). Economic Botany 52: 401-411.
• Piña HH, Montaña C, Mandujano MC. 2007. Fruit abortion in the Chihuahuan Desert endemic cactus Opuntia microdasys Plant Ecology 193: 305-313.
• Primack RB. 1985. Longevity of individual flowers. Annual Review of Ecolgy and Systematics 16: 15-37.
• Potts SG, Biesmeijer JC, Kremen C, Neumann P, Schweiger O, Kunin WE. 2010. Global pollinator declines: trends, impacts and drivers. Trends in Ecology and Evolution 25: 345-353.
• R Core Team. 2020. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/ 14 Jun. 2020.
  » https://www.R-project.org/
• Rasband WS. 2018. ImageJ. Maryland, US National Institutes of Health, Bethesda. https://imagej.nih.gov/ij/ 28 Apr. 2021.
  » https://imagej.nih.gov/ij/
• Razo-León AE, Muñoz-Urias A, Campos-Porras MG, Huerta-Martínez FM, Fierros-López HE. 2022. Flower visitors and efficient pollinators of Opuntia joconostle F.A.C. Weber ex Diguet in Jalisco, Mexico. Bradleya 40: 19-26.
• Reyes-Agüero JA, Aguirre-Rivera JR, Flores-Flores JL. 2005a. Variación morfológica de Opuntia (Cactaceae) en relación con su domesticación en la altiplanicie meridional de México. Interciencia 30: 476-484.
• Reyes-Agüero JA, Aguirre-Rivera JR, Hernández HM. 2005b. Systematic notes and a detailed description of Opuntia ficus-indica (L.) Mill. (Cactaceae). Agrociencia 39: 395-408.
• Reyes-Agüero JA, Aguirre JR, Valiente-Banuet A. 2006. Reproductive biology of Opuntia: A review. Journal of Arid Environments 64: 549-585.
• Rebman JP, Pinkava DJ. 2001. Opuntia Cacti of North America: An overview. Florida Entomologist 84: 474-483.
• Rzedowski J. 1978. Vegetación de México. Mexico, Noriega Editores.
• Sánchez-Echeverría K, Castellanos I, Mendoza-Cuenca L. 2016. Abejas visitantes florales de Opuntia heliabravoana en un gradiente de urbanización. Biológicas 18: 27-34.
• Schlindwein C. 2004. Are oligolectic bees always the most effective pollinator? In: Freitas BM, Pereira JOP (eds.). Bees - Conservation, Rearing and Management for Pollination. Fortaleza, Universidad Federal do Ceará. p. 231-240.
• Sigrist MR, Leme FM, Fabiano VS et al 2021. Chrysophyllum marginatum (Sapotaceae): Generalist pollination and cryptic gynomonoecious. Plant Species Biology 36: 436-449.
• Sipes SD, Tepedino VJ. 2005. Pollen-host specificity and evolutionary patterns of host switching in a clade of specialist bees (Apoidea: Diadasia). Biological Journal of the Linnean Society 86: 487-505.
• Tenorio-Escandón P, Ramírez-Hernández A, Flores J, Juan-Vicedo J, Martínez-Falcón AP. 2022. A systematic review on Opuntia (Cactaceae: Opuntioideae) flower-visiting insects in the world with emphasis on Mexico: implications for diversity conservation. Plants 11: 131.
• Vogler DW, Kalisz S. 2001. Sex among the flowers: The distribution of plant mating systems. Evolution 55: 202-204.
• Waser NM. 1986. Flower constancy: Definition, cause, and measurement. The American Naturalist 127: 593-603.

Supplementary material

The following online material is available for this article:

Figure S1. Study sites of Opuntia atropes were located at the south of Morelos state, in central Mexico.

Figure S2. Floral color variation during anthesis of Opuntia atropes. At the beginning of the anthesis, there was found only one category of color; as flowers age, they turned to more reddish and darker categories of color, according to the Munsell color system.

Publication Dates

• Publication in this collection
  11 Dec 2023
• Date of issue
  2023

History

• Received
  24 May 2023
• Accepted
  30 Sept 2023