Pl. Syst. Evol. (2006) Plant Systematics DOI 10.1007/s00606-006-0493-y and Evolution Printed in The Netherlands On flower visitors and true pollinators: The case of protandrous Heracleum sphondylium L. (Apiaceae) Marcin Zych Warsaw University Botanic Garden, Warsaw, Poland Received April 15, 2005; accepted July 3, 2006 Published online: December 22, 2006 Ó Springer-Verlag 2006 Abstract. Hogweed (Heracleum sphondylium L.), a morphs, pollinator importance, specialization, common European umbellifer, is very variable in generalization. terms of flower and inflorescence morphology. Its flowers are visited by numerous insects, yet little is known about the importance of the particular insect Since the earliest systematic surveys of plants taxa. I observed umbels of two colour morphs pollination systems, the family Apiaceae has (subspecies) of Heracleum sphondylium growing in been regarded as promiscuous (sensu Grant NE Poland, which were visited by more than 108 1949) in terms of pollination biology (Bell insect species during two study seasons. Analysis of 1971, Faegri and van der Pijl 1979, Proctor the insects’ importance suggests that the most et al. 1996). However for the last several efficient pollinators are the medium-sized flies Erio- decades new research based on more detailed zona syrphoides, and Lucilia spp. (Diptera). Bum- blebees Bombus terrestris (Hymenoptera), beetles of studies of the pollinator importance and flower genus Stenurella spp. and Dasytes spp. (Coleoptera) biology has revealed that umbellifers are much and flies Eristalis spp., Meliscaeva cinctella, Phaonia more specialised than previously suggested, angelicae and Thricops nigrifrons also contribute to and their pollination systems may contain pollination of the studied plants, but their efficiency various sophisticated mechanisms overlooked shows considerable seasonal variation. Although the by earlier workers (Zych 2004). Since the dense umbels of the white flowered H. sphondylium works of Bell and Lindsey (Bell and Lindsey subsp. sphondylium are generally more attractive for 1978, Lindsey 1984, Lindsey and Bell 1985), insect visitors than the loose yellowish inflorescences Apiaceae are usually no longer regarded as of H. sphondylium subsp. sibiricum, these taxa do not primitive in terms of pollination biology. seem to attract different sets of the pollinators. For Although some authors (eg. Sheppard 1991, both subspecies, flowers in the staminate phase were Proctor et al. 1996, Rovira et al. 2004) still visited significantly more often than those in the pistillate phase. Some flower visitors visited the consider them ‘promiscuous’, a series of stud- staminate phase only, which suggests they may be ies focused on umbelliferean species from parasites rather than pollinators. different geographic regions show a consider- able degree of specialisation of their pollina- Key words: Heracleum sphondylium, Umbelli- tion systems (Grace and Nelson 1981; Webb ferae, Diptera, pollination, dichogamy, colour 1984; Koul et al. 1986, 1989, 1993; Lamborn M. Zych: Pollination of Heracleum sphondylium and Ollerton 2000; Ne´meth and Sze´kely 2000; Most workers focus on the pollination Schlessman and Graceffa 2002; Zych 2002). biology of economic Apiaceae (Braak and In spite of numerous publications on the Kho 1958; Burgett 1980; Koul et al. 1986, pollination biology of umbellifers, most of the 1993; Warakomska et al. 1986; Koul et al. cited studies, except the works of Lindsey 1989; Koul et al. 1989; Langenberger and (1984), Lamborn and Ollerton (2000), and Davis 2002a, 2002b). From among wild Euro- Rovira and co-workers (2004), are based on pean taxa only three species have received the one-season long observations and lack the attention of researchers – Daucus carota information on long term pollinators activity (Eisikowitch 1980, Lamborn and Ollerton which may be crucial in determining real plant- 2000), Heracleum sphondylium (Corbet 1970, pollinator relationships (Waser et al. 1996, Grace and Nelson 1981, Ellis and Ellis-Adam Johnson and Steiner 2000, Pellmyr 2002). 1994, Zych 2002), and Seseli farrenyi (Rovira Some other aspects of pollination biology et al. 2004). The first was studied because of of the Apiaceae, e.g. insect preferences the Darwin (1888) controversy over the central towards flower sexual phases, have also floret and the second because of its wide received little attention. Although the phe- distribution, importance in supporting local nomenon of dichogamy in this family is well anthophilous fauna (Ellis and Ellis-Adam known, and has been well studied for more 1993), and the very variable morphology of than a century (Foerste and Trelease 1882, the flowers and inflorescences (Gawłowska Knuth 1898, Bell 1971, Webb 1981, Lovett 1961, Sheppard 1991, Zych 2002). The third, Doust and Harper 1980, Lindsey 1982, S. farrenyi, is an endemic species with a very Lovett Doust and Lovett Doust 1982, narrow distribution. Schlessman 1982, Koul et al. 1993, Spalik The case of Heracleum sphondylium espe- and Woodell 1994, Konuma and Yahara cially provides not only useful insight into 1997, Schlessman and Graceffa 2002), sur- pollination biology but also may contribute to prisingly almost none of the works have been the studies of insect preferences and the devoted to the fact that strong dichogamy in evolution of flower morphology (Zych 2002). some cases may cause discrimination of Two of its subspecies (H. sphondylium subsp. sexual phases by insects visitors (Goulson sphondylium and H. sphondylium subsp. sibir- 1999). Among papers devoted to the pollina- icum), formerly considered separate species, tion biology of the umbellifers, only Zych differ in flower and inflorescence morphology. (2002) clearly indicated the sexual phase of The former has a white corolla and zygomor- the observed flowers, and some evidence of phic flowers with elongated petals arranged on discrimination of sexual phases was given by the edge of an umbel, while the latter has Lindsey and Bell (1985), who observed that greenish actinomorphic flowers and umbels flowers in the male phase of plants from the with no such ‘edge effect’ (Gawłowska 1956, American genera Thaspium and Zizia are 1961; Sheppard 1991). visited more frequently by Diptera and some Although, in terms of the number of insect Hymenoptera (e.g. Andrena ziziae) than those visitors, the species is regarded as very ‘pop- in the female phase. This finding however ular’ among European anthopiles (Ellis and was not thoroughly discussed by these Ellis-Adam 1993, 1994), and both taxa seem to authors and passed almost unnoticed by later be pollinated by similar dipterans, preliminary workers, until Schlessman et al. (2004) exper- evidence suggests that their floral characteris- imentally showed that either female stage or tics serve as attractants for different groups of smaller in size umbels of protogynous insects which may create a mechanism of Pseudocymopterus montanus (North American differentiation between the two (Zych 2002). endemic) were less attractive to its floral However this opinion is based only on visitors, mostly solitary bees and flies. short term observations. The present paper M. Zych: Pollination of Heracleum sphondylium addresses the following questions: (1) What is percentage of male flowers is quite variable. In the long-term pollination biology of H. some populations it does not exceed 40% in sphondylium? Does the species have constant tertiary umbels (Wro´blewska 1992), while in others (faithful) pollinators or is there considerable may be close to 100% (Zych, unpubl.). In primary variation in pollinators between years? And (2) umbels hermaphrodite flowers usually constitute Do insect visitors show any preferences regard- 100% (Wro´blewska 1992). Flower colour varies from greenish, greenish-yellow (H. sphondylium ing floral sexual phases? subsp. sibiricum), to white or pinkish (H. sphondy- lium subsp. sphondylium). Material and methods Study site. The observations were conducted in 2000 and 2002 in two sites in the same forest Heracleum sphondylium L. Heracleum sphondylium complex in Wigry National Park (NE Poland). The is a perennial (hemicryptophyte) (Rabotnov 1956, distance between the two populations was ±3 km Sheppard 1991), not a biennial as often stated (e.g. (the same populations were also studied by Zych Brummitt 1968), with cauline leaves arranged in a (2002) in 1998). The population of H. sphondylium rosette producing erect flower stems up to 3 m tall. subsp. sphondylium was situated in the forest It is widespread in forests and woodland clearings, section 119 near the village of Krzywe (N 54°05’ riverbanks, and tall montane-herb grasslands E23°00’), and the population of H. sphondylium (Sheppard 1991). In Poland it is usually described subsp. sibiricum was on the border of the forest as a common meadow species (Gawłowska 1956), sections 133 and 134 (N 54°06’ E 23°03’). Both growing also in thickets, on roadsides and forest grew in similar habitat conditions, along the forest fringes (Rutkowski 1998). Its geographical range road in mixed spruce-pine forest, and consisted of includes most of the European countries, except the similar numbers of individuals. extreme north, parts of the Mediterranean region Field observations. Field observations in both and some Atlantic and Mediterranean islands study years were conducted in July, a peak (Brummitt 1968, Sheppard 1991). flowering month for H. sphondylium in NE Poland. Heracleum sphondylium reproduces mainly by I observed male phase umbels from 4–5 July and seeds; vegetative division of the