American Journal of Botany 101(12): 2148–2159, 2014.
FLIES AS POLLINATORS OF MELITTOPHILOUS SALVIA SPECIES (LAMIACEAE)1 FERHAT CELEP2,6, ZEYNEP ATALAY3, FATIH DIKMEN4, MUSA DOĞAN3, AND REGINE CLASSEN-BOCKHOFF5 2Department of Biology, Polatlı Faculty of Arts and Sciences, Gazi University, 06900 Polatlı, Ankara, Turkey; 3Department of Biological Sciences, Faculty of Arts and Sciences, Middle East Technical University, Ankara, Turkey; 4Department of Biology, Faculty of Science, Istanbul University, Istanbul, Turkey; and 5Institut für Spezielle Botanik, Johannes Gutenberg Universität, Mainz, Germany
• Premise of the study: Floral adaptation to a functional pollinator group does not necessarily mean close specialization to a few pollinator species. For the more than 950 species of Salvia, only bee and bird pollinations are known. Restriction to these pollinators is mainly due to the specific flower construction (lever mechanism). Nevertheless, it has been repeatedly suggested that Salvia flowers might also be pollinated by flies. Are flies able to handle the lever mechanism? Are they functionally equivalent pollinators? In this study, we compared and quantified pollen transfer by bees and flies to test whether flies are true pollinators in Salvia. • Methods: We identified pollinators using field observations and photos. Video documentation of the visitation rate and the site of pollen placement on the pollinator body, morphometric measurements, quantification of pollen placement, pollen load, handling time, and stigma contact ratio were analyzed. • Key results: Field investigations revealed that 19 insect species pollinated S. virgata and four pollinated S. verticillata, including 16 bee species from seven genera of the Apidae and three fly species from three genera of the Nemestrinidae and Tabanidae. • Conclusions: Flies have been found to be pollinators in primarily bee-pollinated Salvia species. This result demonstrates the potential of a given “melittoid” flower construction to broaden the range of pollinators to guarantee successful pollination and seed production. Though bees, particularly Bombus terrestris, were more efficient than flies, the study shows that flies significantly contribute to pollen transfer in Salvia. Key words: Apidae; bee pollination; fly pollination; Lamiaceae; morphometry; Nemestrinidae; pollinator efficiency; Salvia; Tabanidae; Turkey.
The mutualistic interaction between flowers and pollinators often favors the evolution of floral syndromes, i.e., specific sets of floral traits, each associated with a specific functional pollinator group (sensu Vogel, 1954; Fenster et al., 2004; Johnson, 2010). On the basis of floral traits, pollinators may be predicted with some degree of accuracy (Armbruster et al., 2011). However, such predictions should be treated with caution (Ollerton et al., 2007). Flowers need not be specialized to a certain pollinator group; assumed adaptations may actually be misinterpretated or over-interpreted (Ollerton et al., 2009a). In any case, 1 Manuscript received 25 September 2014; revision accepted 10 October 2014. We thank Prof. Dr. Abdullah Hasbenli (Gazi University, Ankara, Turkey) for confirming fly identifications. We are grateful to Mainz University (Germany) designers Doris Franke and Anne Korek for preparing the illustrations, to Markus Jerominek (Mainz University) for checking box plots, to Daniel Drucker for editing the English (Coral Springs, FL, USA and Merton College, Oxford University) and to Ass.Prof. Dr. Kubilay Yildirim (Gazi Osmanpașa University, Tokat) for help with statistical analyses. Comments by Prof. Dr. Stephen G. Weller (Irvine, California, USA), Dr. Petra Wester (Düsseldorf, Germany), and anonymous reviewers improved the manuscript. 6 Author for correspondence (e-mail: [email protected]); present address: Department of Biology, Polatli Faculty of Arts and Sciences, Gazi University, 06900 Polatli, Ankara, Turkey
doi:10.3732/ajb.1400422
field investigations show how well the observed pollinators fit expectations based on syndromes. Of the more than 950 species of Salvia L. (Harley et al., 2004), the floral traits of most species are consistent with the bee-pollination syndrome (Hildebrand, 1865; Dieringer et al., 1991), i.e., with bilabiate flowers of blue-violet or yellowish colors, nectar presented deep in the flower tube, pollen sacs and stigma located below the upper lip, and a landing place provided by the lower lip (Wester and Claßen-Bockhoff, 2011). This construction is interpreted as an adaptation to the body proportions, sensory perception, and foraging behavior of bees (Vogel, 1954; Fenster et al., 2004). At the same time, it protects the flower against extensive pollen collection by bees, thereby saving pollen for the plant’s own reproductive success (Westerkamp and Claßen-Bockhoff, 2007). In Salvia, the pollination process is mediated by a staminal lever mechanism (Claßen-Bockhoff et al., 2003). Each connective of the two stamens is widened and forms two lever arms that are movable around the filament tip. The upper connective arm, usually hidden below the upper lip, bears two pollen sacs (fertile theca). The lower connective arm is often sterile and placed in the middle of the flower entrance, restricting access to nectar. A pollinator searching for nectar pushes this barrier back, lowers the upper connective arm, and has pollen grains loaded on its back (Claßen-Bockhoff et al., 2003, 2004). When the insect visits a second flower, the pollen load is transferred to the stigma, which either protrudes from the flower being
American Journal of Botany 101(12): 2148–2159, 2014; http://www.amjbot.org/ © 2014 Botanical Society of America
2148
December 2014]
CELEP ET AL.—FLY POLLINATION IN SALVIA
touched first by the pollinator (approach herkogamy sensu Webb and Lloyd, 1986) or takes up the position of the faded anthers in older flowers. Phenotypic adaptation to bees does not mean that other insect groups are completely excluded from the process of pollen transfer. It has been repeatedly suggested that Salvia flowers might be pollinated by butterflies or long-tongued flies (Sprengel, 1793; Hildebrand, 1865; Knuth, 1899; Müller, 1921; Visco and Capon, 1970; Faegri and van der Pijl, 1971; Read, 1983; Grases and Ramirez, 1998; Potgieter and Edwards, 2001; Fenster et al., 2004; Wester and ClaßenBockhoff, 2011). Though butterflies often visit Salvia flowers, they have not been observed moving the lever mechanism (see Wester and Claßen-Bockhoff, 2006, 2011, and references cited). Similarly, there are several observations of flies visiting Salvia, but without confirmation of pollination. Flies (Nemestrinus caucasicus Fischer) were recorded feeding on flowers of Salvia tesquicola Klokov & Pobed in southern Russia (Narchuk, 2007). Zamora-Carrillo et al. (2011) observed different members of Syrphidae visiting, but not effectively pollinating S. bogotensis Benth. Fenster et al. (2004) published a photograph showing Raphiomidas Osten Sacken, a member of the Apioceridae (ca. 18 mm tongue length) visiting Salvia mohavensis Greene. Cranmer et al. (2012) reported that Syrphidae were regular visitors to Salvia pratensis L. but probably not pollinators. On the basis of floral traits, Potgieter and Edwards (2001) expected long-tongued flies to be pollinators of the South African Salvia scabra L. and Salvia repens Burchell ex Benth. Wester and Claßen-Bockhoff (2011) assumed that butterflies and/or flies pollinated New World species (S. whitehousei Alziar, members of sect. Audibertia (Benth.) Epling and sect. Echinosphace Benth), and observed a fly visiting the flowers of S. greatae Brandegee. These observations and assumptions indicate that insect pollination in Salvia may be more diverse than generally expected. Flies are a highly diverse group of pollinators with respect to the morphology of their mouthparts and motivation to visit flowers (Vogel, 1954). This diversity is also represented by the many families among them, including the flower visitors Syrphidae, Bombyliidae, Tabanidae, Tachinidae, Nemestrinidae, and Muscidae (Kugler, 1970). According to Ssymank et al. (2008), more than 70 families of Diptera regularly visit at least 555 flowering plant species. In accordance with their motivation, body size and proboscis length (e.g., Vogel, 1954; Kugler, 1970; Goldblatt and Manning, 2000; Anderson et al., 2005), flies either prefer flowers that mimic substrates for egg-laying (carrion flowers), flask-shaped flowers that temporarily trap pollinators (e.g., Ceropegia and Aristolochia, Sakai, 2002; Ollerton et al., 2009b), simple flowers with easily accessible nectar and pollen (e.g., Apiaceae, Euphorbiaceae; see Knuth, 1899; Kugler, 1970; Niemirski and Zych, 2011), or flowers with medium to long tubes predominantly associated to bees and/or butterflies (Goldblatt et al., 1996; Goldblatt and Manning, 2000). Pollinators differ in their performance. To quantify their contribution to a plant’s reproductive success (fruit and seed production) pollination efficiency and effectiveness are distinguished by Ne`eman et al. (2010). Here we consider “efficiency” as the extent to which pollen deposited on stigmas leads to full seed set. “Effectiveness”, in contrast, would be the number of pollen grains delivered by a given pollinator to the stigma of a given flower in a single visit (Ne’eman et al., 2010). The major difference between these two concepts is that pollination effectiveness is measured on a per visit basis.
2149
We report on the pollination ecology of a Salvia assemblage from Central Anatolia, an area that is one of the richest in Salvia species (Hedge, 1982; Kahraman et al., 2012). Ten Salvia species co-occur at the study site (F. Celep et al., unpublished data). While the Salvia species are all visited by several different bee species, flies were additionally observed on S. virgata Jacq. and S. verticillata L. flowers. In this study, we compared and quantified pollen transfer by bees and flies to test whether flies are true pollinators in Salvia.
MATERIALS AND METHODS Study area—Pollination studies were conducted from May to July in both 2010 and 2011 in Ankara, Central Anatolia, Turkey (39°52′56.61″N, 32°46′34.62″E). The study area is a protected steppe area at an altitude of 920– 980 m (Fig. 1A, 2A). It extends over 10 km and is partly urbanized (METU campus area). The vegetation is dominated by herbaceous members of Asteraceae, Fabaceae, Boraginaceae, Apiaceae, and Lamiaceae and planted trees of Pinus nigra J.F. Arnold and Cedrus libani A. Rich. Salvia species—Ten Salvia species co-occur in the area (S. absconditiflora Greuter & Burdet, S. aethiopis L., S. candidissima Vahl, S. cyanescens Boiss. & Balansa, S. hypargeia Fisch. & Mey., S. russellii Benth., S. tchihatcheffii Boiss., S. verticillata, S. virgata, S. viridis L.). Visitation by flies was only observed on S. virgata and S. verticillata (F. Celep et al., unpublished data). Salvia virgata (Fig. 1A) is a perennial herb widely distributed in Eurasia (Hedge, 1982). At the study area, the species is common. We investigated two populations 2 km from each other. Flowers are usually violet blue to lilac. The flower has a falcate upper lip and spoon-shaped lower lever arms (Fig. 1B) guiding the pollinator’s proboscis to the nectar (as seen in S. pratensis L., Reith et al., 2007). The plant provides an attractive food source for pollinators because most individuals flower simultaneously. In 2010 and 2011, flowering time started in the second week of June and lasted until the end of July, with the main flowering peak from mid June to mid July. Flowering time of a single plant lasts 4 wk or longer. The individual plants are erect, (10–)40–120 cm high, and usually produce 1500 or more flowers. Salvia verticillata (Fig. 2A, B) is a perennial herb that is widely distributed within Eurasia (Hedge, 1982). It is less abundant than S. virgata in the study area. Two populations (2 km apart) were investigated in 2011. Flowering started the first week of June and lasted until the end of July, with flowering peaks from mid-June to mid-July. The plants have erect, 10–80 cm long stems. Inflorescence whorls (verticillasters) are densely crowded with 8–40 flowers each. Flowers are usually violet blue (Fig. 2C), lilac and rarely white. A V-shaped, hairy ring is present at the base of the corolla tube. Salvia verticillata lacks the lever mechanism due to the reduction of the lower lever arm. Instead, pollinators have to push back the movable upper lip to gain access to nectar (Hildebrand, 1865; Correns, 1891; Claßen-Bockhoff et al., 2004). Pollinator observation—Insect visitors were observed for 80 h, covering the entire flowering period, including all daylight hours and weather conditions, to estimate the visitation rate. An insect was only classified as a pollinator when it came into contact with both pollen sacs and a stigmatic surface. Observations were documented with photographs and videos from a Canon G10 and a Canon 50D digital camera, as well as a Sanyo TH1 video camera. Based on 42 videos covering 85 min 11 s (S. virgata 62 min 45 s, S. verticillata 22 min 26 s), handling time of the main visitors was calculated. For this purpose, 114 bees and flies (91 from S. virgata and 23 from S. verticillata, Table 1) were watched in slow motion mode. They visited a total of 610 flowers (S. virgata 407 flowers, S. verticillata 203 flowers, Table 1). The videos were also used to reconstruct the percentage of legitimate manipulations of the lever mechanism (defined as pollen placement) and the percentage of stigma contact of the main pollinator species. Stigma contact means that the pollinator touches the stigmatic surface with the pollen bearing part of its body so that pollen can be transferred to the stigma. The site of pollen placement on the insect’s body was easily recognizable by the yellow color of the pollen grains and the increase of the yellow spots on the insects’ bodies after repeated flower visitation. Over 100 captured bees and flies, which were observed as flower visitors before, were examined with a
2150
AMERICAN JOURNAL OF BOTANY
[Vol. 101
December 2014]
CELEP ET AL.—FLY POLLINATION IN SALVIA
2151
Fig. 2. Salvia verticillata and insect visitors. (A) Natural habitat in Ankara, Turkey. (B) Inflorescence structure and flowering sequence. (C) Flower shape of S. verticillata. (D, E) Apis mellifera, pollen loaded on anterior head part. (F, G) Bombus terrestris, pollen placement on anterior head part. (H, I) Pangonius pyritosus, pollen on anterior head part. (J) Bombylius fulvescens hovering in front of a flower (not confirmed as pollinator). Scale bars = (B–J) 5 mm.
dissecting microscope to verify the site of pollen placement. Pollen grains were removed from the insect's body and compared with pollen grains from S. verticillata and S. virgata using a light microscope. To ensure that pollen grains belonged to these species, the size, shape, and aperture type of pollen grains were compared with the result of a recent palynological study (Özler et al., 2011). For quantifying pollen load, pollen grains deposited on 31 (S. virgata) and 18 (S. verticillata) main insect visitors were counted using a Leica binocular stereomicroscope (Table 1). Insects were identified by F. Dikmen. Fly identification was confirmed by A. Hasbenli, Ankara. Voucher specimens of plants (see Appendix 1) and insects were deposited at the Department of Biological Sciences, METU, Ankara.
In this study, the focus is given to pollinator efficiency (defined in the introduction). Single visit experiments allowing quantification and comparison of pollinator effectiveness (see Ne’eman et al., 2010) were not conducted. Morphometric measurements—For testing the morphological matching between flowers and floral visitors, 10 floral (Appendix S1, see Supplemental Data with the online version of this article) and 10 insect morphological traits (Appendix S2, see online Supplemental Data) were measured using digital calipers and a Leica DM1000 light microscope. We measured 30 randomly selected flowers in both Salvia species and between 1–12 individuals of each
←
Fig. 1. Salvia virgata and its pollinators. (A) Inflorescence structure and flowering sequence. (B) Flower in longitudinal section (ovary not shown). (C–E) main bee pollinators. (C) Apis mellifera, pollen placement and stigma contact on posterior thorax parts. (D) Anthophora aestivalis, pollen load on anterior part of thorax. (E) Bombus terrestris, pollen on anterior thorax. (F-J) Fly pollinators. (F) Nemestrinus reticulatus, stigma touches anterior thorax. (G) Nemestrinus reticulatus, pollen load on anterior thorax and head. (H, I) Pangonius pyritosus, pollen on head and anterior thorax. (J) Neorhynchocephalus tauscheri, pollen load on head. (K) Nemestrinus caucasicus observed on S. virgata. (L) Nemestrinus reticulatus sucking nectar from Anchusa sp. (Boraginaceae). Scale bars = (A) 10 mm; (B–L) 5 mm.
2152 TABLE 1.
[Vol. 101
AMERICAN JOURNAL OF BOTANY Quantification of bee and fly (*) pollination in (A) Salvia virgata and (B) S. verticillata.
No.pollinators
Pollinator
Visitation frequency
Pollen placement site (see Fig. 3F)
Estimation of pollen load
No.. insects
Handling time ± SD (s)
Ni/Nf
No.flowers
Lm(%)
Sc(%)
3 7
1.74 ± 1.1 0.90 ± 0.4
5/19 43/111
19 111
89 92
16 72
6
1.20 ± 0.4
11/99
99
81
15
3
4.46 ± 1.6
9/29
A
70
24
2
0.95 ± 0.4
15/65
65
90
21
2
1.14 ± 0.5
1/9
na
—
—
3
0.99 ± 0.2
2/13
na
—
—
4
1.16 ± 0.5
2/19
19
80
10
— — — — —
— — — — —
— — — — —
— — — — —
— — — — —
— — — — —
— — 1
— — 0.93 ± 0.5
— — 4/43
— — 43
— — 57
— — 14
A) Salvia virgata 1 2
Apis mellifera Bombus terrestris
+++ +++
g1+g2 f2+g1
3
*Nemestrinus reticulatus
+++
f1+f2+g1
4
*Pangonius pyritosus
++
f1+f2+g1
5
Anthophora aestivalis
++
f1+f2+g1
6
Anthophora pubescens
++
f2+g1
7
Anthophora robusta
++
f1+f2+g1
8
Eucera cf. pollinosa
++
f1+f2+g1
9 10 11 12 13 14 15 16 17
Anthophora sp.4 Anthophora sp.5 Bombus lucorum Bombus niveatus Eucera sp.1 Eucera sp.2 Eucera sp.3 Megachile pilicrus *Neorhynchocephalus tauscheri
+ + + + + + + + +
f2+g1 f2+g1 f2+g1 f2+g1 g1+g2 g1+g2 f2+g1 f2+g1+g2 f2+g1
thorax: 50–200 head: 20–200 thorax: 20–100 head: 20–100 thorax: 100–200 head: 10–50 thorax: 50–200 head: ca. 20 thorax: ca. 20 head: 50–200 thorax: 20–200 head: 20–50 thorax: ca. 20 head: ca. 20 thorax: 20 — — — — — — — — head: ca. 20
18 19
Nomada sp. Rhodanthidium buteum
+ +
g1+g2 f2+g1
thorax: ca. 20 — —
— —
— —
— —
— —
— —
— —
B) Salvlia verticillata 1 2 3 4
Apis mellifera Bombus terrestris *Pangonius pyritosus Bombus niveatus
+++ +++ ++ +
f1+f2 f1+f2 f1 f1+f2
head: 0–30 head: 20–200 head: 0–10 head: ca. 200
9 3 5 1
1.71 ± 0.7 1.26 ± 0.4 2.26 ± 0.8 0.83 ± 0.2
10/93 4/36 8/62 1/12
— — — —
— — — —
— — — —
Notes: Pluses indicate pollinator visiting frequency from high (+++: observed continuously at plant), medium (++, 5–10 times/observation period) to low (+,
FLIES AS POLLINATORS OF MELITTOPHILOUS SALVIA SPECIES (LAMIACEAE)1 FERHAT CELEP2,6, ZEYNEP ATALAY3, FATIH DIKMEN4, MUSA DOĞAN3, AND REGINE CLASSEN-BOCKHOFF5 2Department of Biology, Polatlı Faculty of Arts and Sciences, Gazi University, 06900 Polatlı, Ankara, Turkey; 3Department of Biological Sciences, Faculty of Arts and Sciences, Middle East Technical University, Ankara, Turkey; 4Department of Biology, Faculty of Science, Istanbul University, Istanbul, Turkey; and 5Institut für Spezielle Botanik, Johannes Gutenberg Universität, Mainz, Germany
• Premise of the study: Floral adaptation to a functional pollinator group does not necessarily mean close specialization to a few pollinator species. For the more than 950 species of Salvia, only bee and bird pollinations are known. Restriction to these pollinators is mainly due to the specific flower construction (lever mechanism). Nevertheless, it has been repeatedly suggested that Salvia flowers might also be pollinated by flies. Are flies able to handle the lever mechanism? Are they functionally equivalent pollinators? In this study, we compared and quantified pollen transfer by bees and flies to test whether flies are true pollinators in Salvia. • Methods: We identified pollinators using field observations and photos. Video documentation of the visitation rate and the site of pollen placement on the pollinator body, morphometric measurements, quantification of pollen placement, pollen load, handling time, and stigma contact ratio were analyzed. • Key results: Field investigations revealed that 19 insect species pollinated S. virgata and four pollinated S. verticillata, including 16 bee species from seven genera of the Apidae and three fly species from three genera of the Nemestrinidae and Tabanidae. • Conclusions: Flies have been found to be pollinators in primarily bee-pollinated Salvia species. This result demonstrates the potential of a given “melittoid” flower construction to broaden the range of pollinators to guarantee successful pollination and seed production. Though bees, particularly Bombus terrestris, were more efficient than flies, the study shows that flies significantly contribute to pollen transfer in Salvia. Key words: Apidae; bee pollination; fly pollination; Lamiaceae; morphometry; Nemestrinidae; pollinator efficiency; Salvia; Tabanidae; Turkey.
The mutualistic interaction between flowers and pollinators often favors the evolution of floral syndromes, i.e., specific sets of floral traits, each associated with a specific functional pollinator group (sensu Vogel, 1954; Fenster et al., 2004; Johnson, 2010). On the basis of floral traits, pollinators may be predicted with some degree of accuracy (Armbruster et al., 2011). However, such predictions should be treated with caution (Ollerton et al., 2007). Flowers need not be specialized to a certain pollinator group; assumed adaptations may actually be misinterpretated or over-interpreted (Ollerton et al., 2009a). In any case, 1 Manuscript received 25 September 2014; revision accepted 10 October 2014. We thank Prof. Dr. Abdullah Hasbenli (Gazi University, Ankara, Turkey) for confirming fly identifications. We are grateful to Mainz University (Germany) designers Doris Franke and Anne Korek for preparing the illustrations, to Markus Jerominek (Mainz University) for checking box plots, to Daniel Drucker for editing the English (Coral Springs, FL, USA and Merton College, Oxford University) and to Ass.Prof. Dr. Kubilay Yildirim (Gazi Osmanpașa University, Tokat) for help with statistical analyses. Comments by Prof. Dr. Stephen G. Weller (Irvine, California, USA), Dr. Petra Wester (Düsseldorf, Germany), and anonymous reviewers improved the manuscript. 6 Author for correspondence (e-mail: [email protected]); present address: Department of Biology, Polatli Faculty of Arts and Sciences, Gazi University, 06900 Polatli, Ankara, Turkey
doi:10.3732/ajb.1400422
field investigations show how well the observed pollinators fit expectations based on syndromes. Of the more than 950 species of Salvia L. (Harley et al., 2004), the floral traits of most species are consistent with the bee-pollination syndrome (Hildebrand, 1865; Dieringer et al., 1991), i.e., with bilabiate flowers of blue-violet or yellowish colors, nectar presented deep in the flower tube, pollen sacs and stigma located below the upper lip, and a landing place provided by the lower lip (Wester and Claßen-Bockhoff, 2011). This construction is interpreted as an adaptation to the body proportions, sensory perception, and foraging behavior of bees (Vogel, 1954; Fenster et al., 2004). At the same time, it protects the flower against extensive pollen collection by bees, thereby saving pollen for the plant’s own reproductive success (Westerkamp and Claßen-Bockhoff, 2007). In Salvia, the pollination process is mediated by a staminal lever mechanism (Claßen-Bockhoff et al., 2003). Each connective of the two stamens is widened and forms two lever arms that are movable around the filament tip. The upper connective arm, usually hidden below the upper lip, bears two pollen sacs (fertile theca). The lower connective arm is often sterile and placed in the middle of the flower entrance, restricting access to nectar. A pollinator searching for nectar pushes this barrier back, lowers the upper connective arm, and has pollen grains loaded on its back (Claßen-Bockhoff et al., 2003, 2004). When the insect visits a second flower, the pollen load is transferred to the stigma, which either protrudes from the flower being
American Journal of Botany 101(12): 2148–2159, 2014; http://www.amjbot.org/ © 2014 Botanical Society of America
2148
December 2014]
CELEP ET AL.—FLY POLLINATION IN SALVIA
touched first by the pollinator (approach herkogamy sensu Webb and Lloyd, 1986) or takes up the position of the faded anthers in older flowers. Phenotypic adaptation to bees does not mean that other insect groups are completely excluded from the process of pollen transfer. It has been repeatedly suggested that Salvia flowers might be pollinated by butterflies or long-tongued flies (Sprengel, 1793; Hildebrand, 1865; Knuth, 1899; Müller, 1921; Visco and Capon, 1970; Faegri and van der Pijl, 1971; Read, 1983; Grases and Ramirez, 1998; Potgieter and Edwards, 2001; Fenster et al., 2004; Wester and ClaßenBockhoff, 2011). Though butterflies often visit Salvia flowers, they have not been observed moving the lever mechanism (see Wester and Claßen-Bockhoff, 2006, 2011, and references cited). Similarly, there are several observations of flies visiting Salvia, but without confirmation of pollination. Flies (Nemestrinus caucasicus Fischer) were recorded feeding on flowers of Salvia tesquicola Klokov & Pobed in southern Russia (Narchuk, 2007). Zamora-Carrillo et al. (2011) observed different members of Syrphidae visiting, but not effectively pollinating S. bogotensis Benth. Fenster et al. (2004) published a photograph showing Raphiomidas Osten Sacken, a member of the Apioceridae (ca. 18 mm tongue length) visiting Salvia mohavensis Greene. Cranmer et al. (2012) reported that Syrphidae were regular visitors to Salvia pratensis L. but probably not pollinators. On the basis of floral traits, Potgieter and Edwards (2001) expected long-tongued flies to be pollinators of the South African Salvia scabra L. and Salvia repens Burchell ex Benth. Wester and Claßen-Bockhoff (2011) assumed that butterflies and/or flies pollinated New World species (S. whitehousei Alziar, members of sect. Audibertia (Benth.) Epling and sect. Echinosphace Benth), and observed a fly visiting the flowers of S. greatae Brandegee. These observations and assumptions indicate that insect pollination in Salvia may be more diverse than generally expected. Flies are a highly diverse group of pollinators with respect to the morphology of their mouthparts and motivation to visit flowers (Vogel, 1954). This diversity is also represented by the many families among them, including the flower visitors Syrphidae, Bombyliidae, Tabanidae, Tachinidae, Nemestrinidae, and Muscidae (Kugler, 1970). According to Ssymank et al. (2008), more than 70 families of Diptera regularly visit at least 555 flowering plant species. In accordance with their motivation, body size and proboscis length (e.g., Vogel, 1954; Kugler, 1970; Goldblatt and Manning, 2000; Anderson et al., 2005), flies either prefer flowers that mimic substrates for egg-laying (carrion flowers), flask-shaped flowers that temporarily trap pollinators (e.g., Ceropegia and Aristolochia, Sakai, 2002; Ollerton et al., 2009b), simple flowers with easily accessible nectar and pollen (e.g., Apiaceae, Euphorbiaceae; see Knuth, 1899; Kugler, 1970; Niemirski and Zych, 2011), or flowers with medium to long tubes predominantly associated to bees and/or butterflies (Goldblatt et al., 1996; Goldblatt and Manning, 2000). Pollinators differ in their performance. To quantify their contribution to a plant’s reproductive success (fruit and seed production) pollination efficiency and effectiveness are distinguished by Ne`eman et al. (2010). Here we consider “efficiency” as the extent to which pollen deposited on stigmas leads to full seed set. “Effectiveness”, in contrast, would be the number of pollen grains delivered by a given pollinator to the stigma of a given flower in a single visit (Ne’eman et al., 2010). The major difference between these two concepts is that pollination effectiveness is measured on a per visit basis.
2149
We report on the pollination ecology of a Salvia assemblage from Central Anatolia, an area that is one of the richest in Salvia species (Hedge, 1982; Kahraman et al., 2012). Ten Salvia species co-occur at the study site (F. Celep et al., unpublished data). While the Salvia species are all visited by several different bee species, flies were additionally observed on S. virgata Jacq. and S. verticillata L. flowers. In this study, we compared and quantified pollen transfer by bees and flies to test whether flies are true pollinators in Salvia.
MATERIALS AND METHODS Study area—Pollination studies were conducted from May to July in both 2010 and 2011 in Ankara, Central Anatolia, Turkey (39°52′56.61″N, 32°46′34.62″E). The study area is a protected steppe area at an altitude of 920– 980 m (Fig. 1A, 2A). It extends over 10 km and is partly urbanized (METU campus area). The vegetation is dominated by herbaceous members of Asteraceae, Fabaceae, Boraginaceae, Apiaceae, and Lamiaceae and planted trees of Pinus nigra J.F. Arnold and Cedrus libani A. Rich. Salvia species—Ten Salvia species co-occur in the area (S. absconditiflora Greuter & Burdet, S. aethiopis L., S. candidissima Vahl, S. cyanescens Boiss. & Balansa, S. hypargeia Fisch. & Mey., S. russellii Benth., S. tchihatcheffii Boiss., S. verticillata, S. virgata, S. viridis L.). Visitation by flies was only observed on S. virgata and S. verticillata (F. Celep et al., unpublished data). Salvia virgata (Fig. 1A) is a perennial herb widely distributed in Eurasia (Hedge, 1982). At the study area, the species is common. We investigated two populations 2 km from each other. Flowers are usually violet blue to lilac. The flower has a falcate upper lip and spoon-shaped lower lever arms (Fig. 1B) guiding the pollinator’s proboscis to the nectar (as seen in S. pratensis L., Reith et al., 2007). The plant provides an attractive food source for pollinators because most individuals flower simultaneously. In 2010 and 2011, flowering time started in the second week of June and lasted until the end of July, with the main flowering peak from mid June to mid July. Flowering time of a single plant lasts 4 wk or longer. The individual plants are erect, (10–)40–120 cm high, and usually produce 1500 or more flowers. Salvia verticillata (Fig. 2A, B) is a perennial herb that is widely distributed within Eurasia (Hedge, 1982). It is less abundant than S. virgata in the study area. Two populations (2 km apart) were investigated in 2011. Flowering started the first week of June and lasted until the end of July, with flowering peaks from mid-June to mid-July. The plants have erect, 10–80 cm long stems. Inflorescence whorls (verticillasters) are densely crowded with 8–40 flowers each. Flowers are usually violet blue (Fig. 2C), lilac and rarely white. A V-shaped, hairy ring is present at the base of the corolla tube. Salvia verticillata lacks the lever mechanism due to the reduction of the lower lever arm. Instead, pollinators have to push back the movable upper lip to gain access to nectar (Hildebrand, 1865; Correns, 1891; Claßen-Bockhoff et al., 2004). Pollinator observation—Insect visitors were observed for 80 h, covering the entire flowering period, including all daylight hours and weather conditions, to estimate the visitation rate. An insect was only classified as a pollinator when it came into contact with both pollen sacs and a stigmatic surface. Observations were documented with photographs and videos from a Canon G10 and a Canon 50D digital camera, as well as a Sanyo TH1 video camera. Based on 42 videos covering 85 min 11 s (S. virgata 62 min 45 s, S. verticillata 22 min 26 s), handling time of the main visitors was calculated. For this purpose, 114 bees and flies (91 from S. virgata and 23 from S. verticillata, Table 1) were watched in slow motion mode. They visited a total of 610 flowers (S. virgata 407 flowers, S. verticillata 203 flowers, Table 1). The videos were also used to reconstruct the percentage of legitimate manipulations of the lever mechanism (defined as pollen placement) and the percentage of stigma contact of the main pollinator species. Stigma contact means that the pollinator touches the stigmatic surface with the pollen bearing part of its body so that pollen can be transferred to the stigma. The site of pollen placement on the insect’s body was easily recognizable by the yellow color of the pollen grains and the increase of the yellow spots on the insects’ bodies after repeated flower visitation. Over 100 captured bees and flies, which were observed as flower visitors before, were examined with a
2150
AMERICAN JOURNAL OF BOTANY
[Vol. 101
December 2014]
CELEP ET AL.—FLY POLLINATION IN SALVIA
2151
Fig. 2. Salvia verticillata and insect visitors. (A) Natural habitat in Ankara, Turkey. (B) Inflorescence structure and flowering sequence. (C) Flower shape of S. verticillata. (D, E) Apis mellifera, pollen loaded on anterior head part. (F, G) Bombus terrestris, pollen placement on anterior head part. (H, I) Pangonius pyritosus, pollen on anterior head part. (J) Bombylius fulvescens hovering in front of a flower (not confirmed as pollinator). Scale bars = (B–J) 5 mm.
dissecting microscope to verify the site of pollen placement. Pollen grains were removed from the insect's body and compared with pollen grains from S. verticillata and S. virgata using a light microscope. To ensure that pollen grains belonged to these species, the size, shape, and aperture type of pollen grains were compared with the result of a recent palynological study (Özler et al., 2011). For quantifying pollen load, pollen grains deposited on 31 (S. virgata) and 18 (S. verticillata) main insect visitors were counted using a Leica binocular stereomicroscope (Table 1). Insects were identified by F. Dikmen. Fly identification was confirmed by A. Hasbenli, Ankara. Voucher specimens of plants (see Appendix 1) and insects were deposited at the Department of Biological Sciences, METU, Ankara.
In this study, the focus is given to pollinator efficiency (defined in the introduction). Single visit experiments allowing quantification and comparison of pollinator effectiveness (see Ne’eman et al., 2010) were not conducted. Morphometric measurements—For testing the morphological matching between flowers and floral visitors, 10 floral (Appendix S1, see Supplemental Data with the online version of this article) and 10 insect morphological traits (Appendix S2, see online Supplemental Data) were measured using digital calipers and a Leica DM1000 light microscope. We measured 30 randomly selected flowers in both Salvia species and between 1–12 individuals of each
←
Fig. 1. Salvia virgata and its pollinators. (A) Inflorescence structure and flowering sequence. (B) Flower in longitudinal section (ovary not shown). (C–E) main bee pollinators. (C) Apis mellifera, pollen placement and stigma contact on posterior thorax parts. (D) Anthophora aestivalis, pollen load on anterior part of thorax. (E) Bombus terrestris, pollen on anterior thorax. (F-J) Fly pollinators. (F) Nemestrinus reticulatus, stigma touches anterior thorax. (G) Nemestrinus reticulatus, pollen load on anterior thorax and head. (H, I) Pangonius pyritosus, pollen on head and anterior thorax. (J) Neorhynchocephalus tauscheri, pollen load on head. (K) Nemestrinus caucasicus observed on S. virgata. (L) Nemestrinus reticulatus sucking nectar from Anchusa sp. (Boraginaceae). Scale bars = (A) 10 mm; (B–L) 5 mm.
2152 TABLE 1.
[Vol. 101
AMERICAN JOURNAL OF BOTANY Quantification of bee and fly (*) pollination in (A) Salvia virgata and (B) S. verticillata.
No.pollinators
Pollinator
Visitation frequency
Pollen placement site (see Fig. 3F)
Estimation of pollen load
No.. insects
Handling time ± SD (s)
Ni/Nf
No.flowers
Lm(%)
Sc(%)
3 7
1.74 ± 1.1 0.90 ± 0.4
5/19 43/111
19 111
89 92
16 72
6
1.20 ± 0.4
11/99
99
81
15
3
4.46 ± 1.6
9/29
A
70
24
2
0.95 ± 0.4
15/65
65
90
21
2
1.14 ± 0.5
1/9
na
—
—
3
0.99 ± 0.2
2/13
na
—
—
4
1.16 ± 0.5
2/19
19
80
10
— — — — —
— — — — —
— — — — —
— — — — —
— — — — —
— — — — —
— — 1
— — 0.93 ± 0.5
— — 4/43
— — 43
— — 57
— — 14
A) Salvia virgata 1 2
Apis mellifera Bombus terrestris
+++ +++
g1+g2 f2+g1
3
*Nemestrinus reticulatus
+++
f1+f2+g1
4
*Pangonius pyritosus
++
f1+f2+g1
5
Anthophora aestivalis
++
f1+f2+g1
6
Anthophora pubescens
++
f2+g1
7
Anthophora robusta
++
f1+f2+g1
8
Eucera cf. pollinosa
++
f1+f2+g1
9 10 11 12 13 14 15 16 17
Anthophora sp.4 Anthophora sp.5 Bombus lucorum Bombus niveatus Eucera sp.1 Eucera sp.2 Eucera sp.3 Megachile pilicrus *Neorhynchocephalus tauscheri
+ + + + + + + + +
f2+g1 f2+g1 f2+g1 f2+g1 g1+g2 g1+g2 f2+g1 f2+g1+g2 f2+g1
thorax: 50–200 head: 20–200 thorax: 20–100 head: 20–100 thorax: 100–200 head: 10–50 thorax: 50–200 head: ca. 20 thorax: ca. 20 head: 50–200 thorax: 20–200 head: 20–50 thorax: ca. 20 head: ca. 20 thorax: 20 — — — — — — — — head: ca. 20
18 19
Nomada sp. Rhodanthidium buteum
+ +
g1+g2 f2+g1
thorax: ca. 20 — —
— —
— —
— —
— —
— —
— —
B) Salvlia verticillata 1 2 3 4
Apis mellifera Bombus terrestris *Pangonius pyritosus Bombus niveatus
+++ +++ ++ +
f1+f2 f1+f2 f1 f1+f2
head: 0–30 head: 20–200 head: 0–10 head: ca. 200
9 3 5 1
1.71 ± 0.7 1.26 ± 0.4 2.26 ± 0.8 0.83 ± 0.2
10/93 4/36 8/62 1/12
— — — —
— — — —
— — — —
Notes: Pluses indicate pollinator visiting frequency from high (+++: observed continuously at plant), medium (++, 5–10 times/observation period) to low (+,
