Pollination ecology and breeding system of two Calceolaria species in Chile

Findings

Many angiosperms are exclusively dependent on pollinators for its reproduction (Matallana et al. 2010; Arroyo et al. 2006). However, pollinators sometimes could be erratic and variable in relation to the ecological context (e.g., plant community composition), especially in alpine zones where it is known that pollinators decline in abundance with the increment in altitude (Arroyo and Squeo 1990; Totland 1994). The latest could be critical in specialized pollination system and more over when specialized flowering plants inhabit in sympatry, potentially sharing the specialized floral visitors. In this context, it is expected that plant species develop reproductive strategies to ensure reproduction and/or exhibit some differences in their pollination ecology.

Calceolaria filicaulis Graham (Calceolariaceae) and Calceolaria arachnoidea Clos (Calceolariaceae) are perennial herbs that grow in streams and rivers in the high Andean region of Chile (Ehrhart 2000). Except for some specific localities in central Chile, the two species have ranges that do not overlap (Ehrhart 2000). The flowers are zygomorphic and composed of a small superior lobe that covers the reproductive structures and an inflated lower lobe that contains the oil-secreting gland (Sérsic 2004). Species differ in many morphological traits, especially those related to corolla characteristics (Figure 1). On the one hand, C. filicaulis has a yellow corolla and the oil gland is situated close to its aperture, enabling a better access to pollinators. On the other hand, C. arachnoidea has a purple corolla and the oil gland is hidden in the inferior lobe. Recent studies indicate that the pollination service of this genus is done by a particular group of bees of the genera Centris and Chalepogenus that possess specialized structures for oil collection (Figure 1; Sérsic 2004).

Figure 1

Plants, corolla close up, and principal pollinators of each Calceolaria species. (A) C. filicaulis and Chalepogenus caeruleus. (B) C. arachnoidea and Centris nigerrima.

This study was conducted during the summer season of 2012 at the National Reserve Altos de Lircay (35°36′ S, 71°00′ W) in the Región del Maule, Chile. The sample site is located in the Reserve at 2,200 m a.s.l. and has a Mediterranean-type climate, with most rainfall concentrated in the winter season (di Castri and Hajek 1976). In the study site, C. filicaulis and C. arachnoidea inhabit in sympatry and are distributed in monospecific patches, surrounded by a vegetation matrix that includes Mimulus luteus Linnaeus (Schrophulareaceae), Mimulus cupreus Dombrain (Schrophulareaceae), Hypochoeris acaulis Britton (Asteraceae), Pozoa coriacea Lagasca & Segura (Apiaceae), and Azorella incisa Weddell (Apiaceae).

To determine the breeding system of Calceolaria species, a pollination-controlled experiment was conducted in 120 buds per species. Buds were bagged with a mesh, and each of the following treatments were randomly assigned: (1) control (C), intact flowers exposed to natural pollination; (2) hand cross-pollination (HC), emasculated flowers pollinated with pollen from another plant situated at least 1 m apart; (3) automatic pollination (AP), bagged buds without manipulation; and (4) hand self-pollination (HS), emasculated flowers pollinated by its own pollen. Pollination success was estimated as seed set per plant, and self-incompatibility (ISI) and autogamy (IAS) indexes (Arroyo and Uslar 1993) were estimated as

$$\mathrm{ISI}=\frac{{\mathit{n}}^{\circ }\overline{\mathrm{HS}}}{{\mathit{n}}^{\circ }\overline{\mathrm{HC}}}\mathrm{IAS}=\frac{{\mathit{n}}^{\circ }\overline{\mathrm{HS}}}{{\mathit{n}}^{\circ }\overline{\mathrm{AP}}}$$

where, $\overline{\mathrm{HS}}$, $\overline{\mathrm{HC}}$, and $\overline{\mathrm{AP}}$ represents the average number of seed per fruit on each treatment. The significance of treatment effects was evaluated on log-transformed data by two-way ANOVA. Statistical differences among treatments were evaluated in a post hoc Tukey test in R package version 2.15 (R development Core Team, 2011). To characterize pollination ecology, all flower visitors of both Calceolaria species were recorded. Focal observations of 15 min per plant were realized during six consecutive sunny days between 0900 and 1800 hours (50 h of observation per specie). All floral visitors contacting the reproductive structures of the flowers were captured and taken to the laboratory for taxonomic identification.

The results of the pollination experiment showed that no seeds were produced in the absence of pollinators. The comparison between treatments revealed that there was a significant effect of treatments on seed production for both Calceolaria species (C. filicaulis: F_2,66 = 4.85, P = 0.01; C. arachnoidea: F_2,82 = 28.34, P < 0.0001). Tukey's post hoc comparison showed that for both species, HS produced less seed compared to HC (C. filicaulis: P = 0.01; C. arachnoidea: P < 0.0001). However, the comparison between natural and supplementary pollination treatments showed different results. In C. filicaulis, neither HS (P = 0.15) nor HC (P = 0.27) showed significant differences with control conditions, while in C. arachnoidea, natural pollination produced more seeds per fruit than HS and HC treatments (P < 0.0001). Finally, the controlled hand pollination indicated that C. filicaulis is allogamous (IAS = 0) and self-incompatible (ISI = 0.14), whereas C. arachnoidea is partially autogamous (IAS = 0.21) and self-compatible (ISI = 0.22). Pollinator assemblages of both Calceolaria species were totally composed of hymenopterans. However, Calceolaria species differ in terms of the number of species responsible for pollination (Table 1). For C. arachnoidea, the pollinator assemblage was composed of four species, while for C. filicaulis, just two species were responsible for all pollination service. The overall visit rate (mean ± SE) was lower for C. arachnoidea (0.09 ± 0.07) than for C. filicaulis (0.17 ± 0.12). Nonetheless, the identity of the principal pollinator was different in both species. For C. arachnoidea, the principal pollinator was Centris nigerrima Spinola (Apidae) (86% of total visits) and for C. filicaulis was Chalepogenus caeruleus Friese (Apidae), representing 82% of total visits.

Table 1 Floral visitors of each studied Calceolaria species

In general, our results showed that Calceolaria species are highly dependent on pollinator for reproduction; however, they are able to self-fertilize and showed differences in their reproductive strategies. C. filicaulis is allogamous and self-incompatible, whereas C. arachnoidea is partially autogamous and present an incomplete self-compatibility. The latest seems to be frequent in flowering plants, in particular in specialized pollination systems (Perez et al. 2009), where the evolution of mixed mating strategies has evolved as stable strategy to deal with variable pollinator environments (see review in Goodwillie et al. 2005). In addition, both Calceolaria species were mainly visited by oil-collecting bees, but they differ in their principal pollinator. Globally, this finding suggest that Calceolaria and their oil-collecting bees could represent a highly specialized system, where plant species develop different reproductive strategies possibly in response to their ecological context. The latest could be explained by the frequency, consistency, and efficiency of their pollinators, which are known to decrease with the increase of elevation (Arroyo et al. 1985, 2006). In fact, in the study site, oil-collecting bees are present in low abundance, and they visit different plant species searching for pollen (MM, personal observation); it is known that this can interfere in plant fecundation through pollen interference and/or competition (Kunin 1997).

Despite that Calceolaria species are one of the most diverse genus in Chile (Fuentes et al. 1995), their biology, systematics, and evolution have been recently the subject of detailed research (Molau 1988; Sérsic 2004; Cosacov et al. 2009). Consequently, although this work represents a preliminary view of the breeding system and pollination ecology of Chilean Calceolaria species, this can be a starting point in the study of ecology and even evolution (e.g., reproductive barriers to speciation) of the genus. In this context, more studies are needed to determine proximal and ultimate mechanisms that are determining this finding in one of the most specialized plant-pollinator systems in Chile.