| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
What's this? |
Reproductive Biology |
School of Biological and Conservation Sciences, University of KwaZulu-Natal, P Bag X01, Scottsville, Pietermaritzburg 3209, South Africa
Received for publication April 4, 2006. Accepted for publication February 20, 2007.
ABSTRACT
The African orchid flora has a high proportion of species with long-spurred white flowers. Few data exist to test the prediction that this floral syndrome pattern reflects an important role for hawkmoth pollination in the evolution and ecology of these orchids. The pollination biology of five aerangoid orchid species (Rangaeris amaniensis, Aerangis brachycarpa, A. confusa, A. thomsonii, and A. kotschyana) was investigated in Kenya. Four of these have long spurs (>10 cm) and were pollinated by Agrius convolvuli and Coelonia fulvinotata. Aerangis confusa, which has relatively short spurs (ca. 4 cm), was pollinated by the short-tongued hawkmoths Hippotion celerio and Daphnis nerii. Nectar frequently filled the entire spur in some of the study species, even at anthesis. Sugar concentration of the nectar of four species was found to vary from ca. 1% at the mouth of the spur to 20% at the tip. Gradients were expressed more strongly in species with long, straight spurs. Species with spirally twisted spurs showed both steep and shallow nectar gradients. These gradients, previously unknown in plants, may function as a "sugar trail," enticing long-tongued hawkmoths to probe deeply into spurs without incurring the cost of filling an entire spur with concentrated nectar. In addition, the most concentrated nectar is kept out of reach of short-tongued pollinators.
Key Words: Aerangis • Africa • Agrius convolvuli • hawkmoths • insect proboscis • Orchidaceae • pollination syndrome • Rangaeris • sphingophily
The long proboscides of hawkmoths are typically matched and even exceeded in length by the corolla tubes of flowers pollinated by these insects. Darwin (1862)
famously argued that the long spurs (ca. 32 cm) of Angraecum sesquipedale, a Madagascan orchid species, represents an example of floral specialization for pollination by a long-tongued hawkmoth. This classic example has been widely cited as evidence of coevolution. Subsequent studies have confirmed the importance of deep corolla tubes for effective pollination by hawkmoths, particularly when the head region of the moth acts as the pollen-bearing surface (Nilsson, 1988
; Alexandersson and Johnson, 2002
) Orchids with long-spurred white flowers and crepuscular scent production are common in Africa (Dressler, 1981
). It is likely that many of these are adapted for pollination by hawkmoths because these floral traits are consistent with the general floral syndrome of hawkmoth pollination (cf. Grant and Grant, 1983; Grant, 1985
; Haber and Frankie, 1989
).
In this study we focused on the pollination of aerangoid orchids (Aerangidinae) in eastern Africa. These long-spurred species offer a chance to investigate the ecology of specialized pollination systems in the wild. Orchids in this subtribe tend to have nectariferous spurs, which in many species are over 10 cm in length. No detailed studies have been conducted on the pollination of epiphytic orchids in eastern Africa. Only one account was found in a local natural history journal noting hawkmoths visiting orchids and other flowers in a Nairobi garden (Ng'weno, 1985).
Aerangis with c. 70 species is the largest genus in the Aerangidinae in Africa and Madagascar. At least 21 Aerangis species are found in eastern Africa (Blundell, 1987
; Cribb, 1989
; Agnew and Agnew, 1994
; Stewart, 1996
), indicating that the region represents a center of diversity alongside Madagascar. Like many other orchid genera, the aerangoid orchids have similarities in floral morphology and behavior, suggesting that pollinators play an important role in their evolution and speciation (Dressler, 1980).
The flowers of Aerangis and those of the smaller related genus Rangaeris (consisting of six species, all found in Africa) are borne on spikes from the leaf bases of the monopodial plants. Hot, dry weather followed by intense rains produces abundant flowering in these and other epiphytic orchids in eastern Africa. In some cases, individual plants form large clumps on trees when undisturbed and can produce hundreds of flowers during a season.
The structure of aerangoid orchid flowers restricts access to nectar in long, narrow tubes that project behind the flower from the lip. To access the nectar, a pollinator has to insert its proboscis through the narrow opening to the spur at the base of the orchid's labellum and reach down into the narrow tube. The only insects with mouthparts capable of reaching into such long, narrow tubular flowers are Lepidoptera, especially the long-tongued hawkmoths (Darwin, 1862
; Miller, 1984
; Grant, 1985
; Nilsson et al., 1987
; Johnson, 1995
; Johnson and Liltved, 1997
; Johnson et al., 2002
, 2004
). The proboscides of the hawkmoths (Sphingidae) are amongst the most highly developed of Lepidoptera and adapted to probing and accessing nectar from flowers (Miller, 1997
).
The size, dispersal, and flight patterns of long-tongued hawkmoths combine to make them ideal agents for carrying pollen or pollinaria rapidly and efficiently over long distances between individual plants (Grant, 1985
; Nilsson, 1992
; Johnson, 1995
; Johnson and Liltved, 1997
). Aerangoid orchids in Kenya generally occur singly or in large clumps on isolated large trees and/or on rock faces. Hawkmoths are ideal agents for transferring pollinaria between individual plants in these habitats because they can disperse widely over short periods of time.
The role of hawkmoths as pollinators has been established mostly by capturing specimens and quantifying their pollen loads (Grant, 1985
, 1992
; Nilsson et al., 1987
; Alexandersson and Johnson, 2002
). Angraecum sesquipedale and other long-spurred aerangoid and angraecoid species have been studied a number of times in Madagascar, and various species of hawkmoths have been found to be their pollinators (Nilsson et al., 1987
; Wasserthal, 1997
). However, some studies conducted in Madagascar have been inconclusive due to the difficulty of observing hawkmoth visits to orchids in the wild and have instead focused on collecting hawkmoths by light-trapping and experiments in flight cages (Wasserthal, 1997
). Earlier studies in Madagascar established that only certain long-tongued hawkmoth species were reliable orchid pollinators. This was assessed through an analysis of pollinaria or viscidia borne by light-trapped hawkmoths, as well as through direct observation. These studies also confirmed the tendency for placement of pollinaria on the base of the proboscis, eyes, or frons of hawkmoths (Nilsson et al., 1987
).
The specific aims of the study were to describe and illustrate aspects of the pollination biology of Aerangis and Rangaeris species in Kenya. To establish whether long-tongued hawkmoths are the primary pollinators of long-spurred aerangoid orchids, we sought answers to the following general and specific questions: (1) What are the correlations between hawkmoth proboscis lengths and aerangoid orchid floral spur lengths? (2) What characteristics of long-tongued hawkmoths make them reliable pollinators? (3) What are the properties of the nectar rewards in complex aerangoid orchid flowers?
MATERIALS AND METHODS
The aerangoid orchid species
The floral characteristics, nectar production, flowering phenology, and pollinators of the following five aerangoid orchid species were investigated: Aerangis brachycarpa (A. Richard) Durand & Schinz, Ae. confusa J. Stewart, A. thomsonii (Rolfe) Shlechter, A. kotschyana (Reichenbach f.) Schlechter, and Rangaeris amaniensis (Kraenzlin) Summerhayes. All species bear star-shaped white or off-white flowers with medium length, ca. 4–5 cm (A. confusa), or long, ca. >10 cm (A. brachycarpa, A. thomsonii, A. kotschyana, and R. amaniensis), floral spurs. The spurs of all species studied contain nectar as a reward to pollinators. The spurs project backward from the lip and either hang straight or are spirally twisted (see Table 1, Fig. 1) (Blundell, 1987
; Cribb, 1989
; Agnew and Agnew, 1994
; Stewart, 1996
). The buds of all species studied open in the late afternoon/early evening. The petals are initially held partly forward, and the sepals are initially tinted green in all the Aerangis spp. As the flower opens more fully over a couple of nights, the petals reflex and become brilliant white. The flowers are strongly scented in the evening. The chemical composition of floral fragrance for all of the study species except A. thomsonii is documented by Kaiser (1993)
. The stigmas of all species studied are sticky and located behind a rostellum in a shallow depression within the column.
|
|
Aerangoid orchids in Kenyan forests flower, broadly speaking, at the beginning or in the months following the "long rains." These rains mark the start of the main growing season and the end of the dry season (January–February). Heavy storms accompanied by lighting and thunder are normal, with less heavy showers in May and June. All Aerangis species were in flower for about 1–2 mo between April and August. Aerangis brachycarpa flowered again in November. Rangaeris amaniensis flowered from March to July and again in November (during the "short rains").
Study sites
The populations of orchids studied were distributed throughout Kenya in forest, woodland and forest edge habitats. Forest fragments around Nairobi (038°7' E, 01°4' S) were the main study site for A. brachycarpa, A. confusa, and R. amaniensis. Aerangis thomsonii was investigated primarily in the Aberdares (036°6' E, 00°9' S) and A. kotschyana along the Kenyan coast (040°1' E, 03°3' S) as well as in western Kenya–Uasin Gishu and Trans Nzoia District.
Observations on A. brachycarpa, A. confusa, and R. amaniensis were made around Nairobi (Nairobi Province and northern Kajiado District) in March, April, and May of 2004 and 2005. Aerangis brachycarpa was also observed in the Mukogodo forest north of Mt. Kenya in Laikipia District. Aerangis thomsonii was observed in the Aberdares (Kiambu District) and in Nairobi (Kitisuru) on a number of occasions in April 2004 and July–August 2005. Aerangis kotschyana was observed in western Kenya in May–July 2004 and at the coast (Malindi–Watamu) on a number of occasions in June–August 2005. Studies of flowering periods, fruit set in the wild, and growth habit were made on all plants observed at all sites.
Hawkmoth visitation observations
Approximately 40 h of focused observations were carried out at various sites over 2 years (2004–2005). Flowering clumps of orchids were carefully watched for pollinator activity and behavior at all sites (details of observation times are given in Table 1, Table 6, and Fig. 2). The time of each hawkmoth visit was recorded along with ambient environmental conditions (local weather and cloud cover), number of individual open flowers, number of flowers visited, and the identity of the visitor. Visitation data were recorded in 1-min intervals. Flowers were watched using a number of different methods, primarily direct observation with or without a dimmed torch. Observations were made from ca. 1800 hours until ca. 30 min after the last hawkmoth was observed at all study sites.
|
Floral morphology and nectar volume and concentration
Fully developed flowers were measured on all plants for the five species studied at various sites. Additional measurements were made of specimens in cultivation and of collected plants and herbarium specimens at the East African Herbarium, National Museums of Kenya. Spur lengths and general morphological features were recorded.
Nectar samples were taken from fully open flowers detached in the early evening (ca. 1730–1800 hours) before pollinator activity began. Nectar was extracted from the spurs immediately using graduated microcapillary tubes (5 or 100 µL) by carefully cutting the spurs into regular-length sections (10–20 mm) and drawing the nectar out of each section using capillary action. Volumes from individual spur sections were noted and correlated with the sections of spur length and distance from the mouth of the spur or distance from the top of the nectar column in the spur. Nectar concentrations were ascertained with a pocket refractometer (Bellingham and Stanley, Turnbridge Wells, UK). The relationship between nectar properties (volume and concentration) and distance from the tip of the floral spur was ascertained using linear regression. When interesting trends in nectar concentrations were noticed, further sampling was done on plants that were isolated in the netted cage and greenhouses of the National Museums of Kenya and Kenya Orchid Society to control for factors such as rainfall, exposure to sunlight, and potentially unobserved pollinator visitation that could potentially have affected the nectar volumes and concentrations.
Breeding system and pollination success
To determine whether any of the study species were capable of setting fruit in the absence of pollinator visits, we enclosed a number of spikes in bud on each species in bags constructed of fine netting with drawstrings at the top. A total of eight spikes were bagged and treated, two for each species with the exception of A. kotschyana. On each manipulated spike, two flowers were selfed and two were crossed. The other bagged flowers were left untreated as controls. In addition, plants of A. brachycarpa, A. confusa, A. thomsonii, and R. amaniensis were maintained in isolation in a large netted cage (2.5 x 2.5 m). Fruit set was also assessed in the wild to determine the percentage of flowers pollinated (Table 5). Natural levels of fruit set were ascertained by counting the number of open, fully developed flowers on individual plants or clumps and later recording the number of fruits produced that developed to maturity. This was done at a range of sites encompassing broad biogeographic and local climatic variability.
|
RESULTS
Hawkmoth visitation rates and behavior at flowers
More than 70 individual hawkmoth visitation bouts were recorded during this study. Hawkmoths were observed at all study sites on all the orchid species in flower. Three long-tongued hawkmoth species were observed visiting the long-spurred orchids and one short-tongued moth species was seen on the shorter-spurred species Ae. confusa (Table 2). Hawkmoth visitation took place in a remarkably restricted period of time from 1846 hours (earliest visit recorded) to 1907 hours (latest visit recorded) (Fig. 2). Individual visits lasted between 2 and 5 s. However, hawkmoths were not present on all nights of observation and visitation rates varied greatly between sites, species, and even within species (Fig. 2). Local weather conditions were one limiting factor for visitation by hawkmoths: they were not seen flying in heavy rain or strong winds. During overcast, drizzly conditions, cloudy conditions without rain, and clear evenings, hawkmoths flew about and fed from flowers at all sites studied.
|
When hawkmoths inserted their proboscides into a flower, they dropped down and then flew upwards until their proboscides were fully inserted into the spur and their heads positioned flush with the column (Fig. 4b, c). This component of foraging was rapid, taking place in 1–2 s. Sometimes a moth grasped the flower with its legs while still hovering with its proboscis fully inserted into the spur. After feeding on the available nectar, hawkmoths withdrew by flying out backwards in a half-somersault (Fig. 4b, c), dropping away from the flower and then hovering briefly before either flying away or beginning to probe another flower on the same plant. Typically, 3–7 flowers per inflorescence were probed by hawkmoths.
|
|
When a hawkmoth carrying pollinaria probes another flower, the base of its head comes up flush against the lower inner part of the column. Here the pollinaria are easily transferred onto the sticky stigma, that is a characteristic of all the aerangoid orchids studied. The pollinaria are pressed against the concave surface of the stigma where they adhere due to presence of a sticky mucilage. The pollinia readily detach from the stipe that holds them to the viscidium. If the flower has not been visited and had its pollinaria removed, the moth may remove further pollinaria on exiting. However, not all hawkmoth visits were seen to involve the moth reaching deep into the spur and touching the column. Sometimes the hawkmoths simply probed the entrance of the flower before moving to another or exited after inserting their proboscises just a couple of centimetres into the spur. These visits did not result in the detachment or insertion of pollinaria. In addition, early in the season it was observed that freshly emerged hawkmoths visiting R. amaniensis and A. brachycarpa spent longer periods of time hovering in front of flowers while attempting to locate the nectar. It was also noticed that C. fulvinotata grooms itself before flight in the evenings, and this could account for the low numbers of specimens with pollinaria.
Hawkmoth morphology and proboscis lengths
Hawkmoth proboscis lengths varied slightly among members of the same species. The proboscides of the more common visitors to the long-spurred aerangoid orchids averaged about 10 cm in length (Table 3). A significant difference was found between tongue lengths of the male and female hawkmoths of the hawkmoth C. fulvinotata (males have longer proboscis lengths), the only species for which sufficient samples could be obtained (t test, P < 0.0001). Males were more common and fed more widely from flowers than did females. Males also outnumbered females at light traps at all sites.
|
|
|
Scent production
All five aerangoid orchid species studied produced intense scent every evening throughout their flowering periods. In the evening, generally near 1815 hours, scent production became stronger and by 1845 hours the scent can be detected by humans many meters from the flowers, even when they are hidden among the branches of a tree. Scent production appeared to peak for ca. 30 min from 1900 hours and then remained in evidence throughout the night. The scent produced was sweet and spicy. Aerangis brachycarpa produced a strong, sweet peppery scent, the scent of A. confusa is distinctly jasmine-like, and A. thomsonii produces a sweet vanilla-like scent. Rangaeris amaniensis has a milder, lemon-like sweet scent.
Breeding system and fruit set in the wild
None of the bagged, unmanipulated flowers produced any fruit, indicating that the study species are dependent on pollinators for seed production. The pollinaria remained in place on these bagged spikes and the "cap" covering them dried up a few days before the rest of the flower wilted. Selfed flowers' ovaries began to swell but half of these did not continue to develop in the case of R. amaniensis and A. brachycarpa. On the other two species, A. confusa and A. thomsonii, pods developed and ripened from both selfed and outcrossed flowers. All outcrossed flowers developed pods.
These preliminary breeding system experiments suggest that all of the aerangoid orchids studied are self-compatible, as is typical for most orchids (Agnew, 1986
; Tremblay et al., 2005
). Fruit set in the wild was found to be consistent, if variable, at all sites, with 5–17.5% of all fully developed open flowers developing pods (Table 5). In general, large clumps with more flowers produced more pods than individual plants with fewer flowers (Table 5).
DISCUSSION
The results of this study indicate that long-spurred aerangoid orchids in Kenya have highly specialized pollination systems involving just a few long-tongued hawkmoth species. Only three species of long-tongued hawkmoths—Xanthopan morgani morgani, Coelonia fulvinotata, and Agrius convolvuli—were observed during the 2 yr of this study (Table 2). Of these, only two species—C. fulvinotata and Ag. convolvuli—were common at the main study sites. The most abundant long-tongued hawkmoth across all environments and especially in drier areas was Ag. convolvuli, while C. fulvinotata was the most common long-tongued hawkmoth species encountered in forest and woodland habitats. Carcasson (1976)
mentions another long-tongued species, Callosphingia circe (Fawcett, 1915), that occurs in Kenya, but this species was neither observed nor collected during this study. No data are provided on the exact tongue-length of C. circe other than it being "long and thick" (Carcasson, 1968
, 1976
). Long-tongued and other hawkmoths are abundant at the beginning and after the rainy seasons in East Africa (Pinhey, 1962
; Robertson, 1977
; Kingston and Nummelin, 1998
). This seasonal peaking of adult hawkmoth abundance (Owen, 1969
; Carcasson, 1976
) appears to be tracked by the flowering phenology of the orchids in this study (Table 1).
We have identified 19 orchid species of a total of ca. 240 in Kenya (Stewart, 1996
) with spurs longer than 100 mm. These comprise eight aerangoid orchids (Aerangis brachycarpa, Ae. thomsonii, Ae. kotschyana, Ae. confusa, Ae. somalensis Schlechter, Rangaeris amaniensis, and Tridactyle tricuspis Schlechter), one angraecoid orchid (Angraecum infundibulare Lindley), as well as 10 terrestrial species (Bonatea steudneri Reichenbach f. Durand & Schinz, Habenaria altior Rendle, H. armatissima Reichenbach f., H. attenuata Hooker f., H. cavatibrachia Summerhayes, H. cirrhata Lindley Reichenbach f., H. egregia Summerhayes, H. macrura Kraenzlin, H. macruroides Summerhayes, and H. walleri Reichenbach f.). It is likely that all of these species are specialized for pollination by long-tongued hawkmoths. The proportion of orchid species adapted for pollination by long-tongued hawkmoths in Kenya is comparable with Madagascar, although the absolute number of species adapted for long-tongued hawkmoths is greater in Madagascar. Interestingly, studies on Madagascar have also revealed that long-spurred orchids are pollinated by a small group of hawkmoth pollinators (Nilsson et al., 1987
; Nilsson, 1988
).
Nectar gradients in long-spurred aerangoid orchids
The discovery of nectar sugar concentration gradients in a number of the aerangoid orchid species studied (Fig. 3) raises questions about the mechanism responsible and the possible functional significance of such gradients. It is unlikely that such gradients arise through a purely passive process of settlement of sugars within solution, as we have found no evidence that gradients develop after 12 h within a 25% sucrose solution added to the full length (20 cm) of plastic drinking straws with a 5-mm diameter. However, a column of concentrated sugar solution in the basal section of a narrow diameter tube does not mix readily with an adjoining column of a more dilute solution because of adhesive forces between the liquids and the tube walls that are similarly responsible for capillary action (N. Chetty, University of KwaZulu-Natal, personal communication). The viscous drag in the narrow tube will also contribute to slow mixing. We have verified this by adding a 30% sugar solution to the lower section and water to the upper section of the drinking straws described earlier. After 12 h, samples taken from the upper column of these straws have a sugar concentration of <1%, while samples from the base of the column have a sugar concentration of 30% (S. D. Johnson, unpublished data). Thus it is likely that the concentration gradients observed in the study species arise from secretion of sugar at the tip of the spur or reabsorbtion of sugars from the upper part of the spur, or a combination of both, and are maintained by adhesive forces that limit diffusion.
Nectar gradients were most evident in those species that tend to fill almost the entire spur with nectar. This is an unusual phenomenon in plants and may be affordable only in terms of carbon resources if the nectar sugar concentration was not uniformly concentrated. Having observed the clumsy attempts of hawkmoths to feed from long-spurred orchids, we suggest that filling most of the spur with nectar, such that it is easily encountered by the tip of the proboscis, may entice the moth to continue probing to the tip of the spur, which is essential to ensure removal and deposition of pollinaria. The gradient of increasing sugar concentration is likely to further encourage the pollinator to continue probing further down into the flower, thus leading to effective pollination.
Aerangoid orchids appear to be capable of reabsorbing sugars from unused nectar following pollination (Koopowitz et al., 1998
; Luyt and Johnson, 2002
). Given that spurs of aerangoid orchids are often considerably longer than the tongues of their hawkmoth pollinators (Tables 3, 4; Nilsson et al., 1987
), the most concentrated nectar at the tip of the spur will often be unused. It appears that some plants are able to utilize the reabsorbed sugars from pollinated flowers for subsequent seed production (Luyt and Johnson, 2002
). It is not known with certainty whether aerangoid orchids reabsorb sugars from unpollinated flowers, although such a mechanism would be enormously beneficial as smaller clumps of orchids can flower for several years in succession without being pollinated (D. J. Martins, personal observation).
Although it was impossible to determine with any certainty whether the hawkmoths returned to the same individual plants on subsequent nights, our impression was that hawkmoths become more adept at extracting nectar from flowers when we observed the same plant on many consecutive evenings. Learning behavior in response to flowers has been documented in the diurnal hawkmoth species Macroglossum stellatarum Linnaeus (Kelber, 1996
, 1997
, 2002
; Kelber and Pfaff, 1997
), as well as the nocturnal species Manduca sexta (Goyret and Raguso, 2006
). In butterflies, once a particular species has been learned, there are costs to switching foraging patterns (McNeely and Singer, 2001
; Weiss and Papaj, 2003
). If hawkmoths are repeat visitors to particular plants, selection might favor particularly large rewards if these increase the chances of repeated visitation. Repeated visitation would be beneficial if flowers open in sequence or if hawkmoths take several days to learn to probe the flowers correctly. On the other hand, repeated visitation could lead to higher levels of geitonogamous self-pollination. Marking of individual hawkmoths in future experiments would be a valuable means of establishing whether individuals forage locally on consecutive evenings.
It is interesting that in Aerangis species that are adapted for hawkmoth pollination, two distinct patterns of spur morphology and nectar production are observed. For A. brachycarpa and A. confusa, the spurs are full of nectar, showing a steep and clear gradient from the top of the nectar column that begins close to the spur's mouth to its tip. Also notable in A. brachycarpa and A. confusa is that the spurs are relatively straight, once they project backward from the lip. In contrast, in A. thomsonii and A. kotschyana, the nectar is located only in the basal third/quarter of the spur in some flowers. Additionally, both A. thomsonii and A. kotschyana have spirally twisted spurs. This phenomenon (basal nectar column and twisted spur) has also been observed on Angraecum arachnites Schltr. in Madagascar (Nilsson, 1985).
As long-spurred orchid flowers need to entice hawkmoths to reach all the way into the spur, nectar gradients and spirally twisted spurs could both function toward the same end: delaying hawkmoth probing slightly while encouraging them to reach deeper into the flower. This in turn results in pollinaria detachment and adherence to the "correct" part of the hawkmoth's head/proboscis. Such complex systems indicate an asymmetrical but stable mutualism between hawkmoths and long-spurred orchids. However, as the degree and number of spirals varies greatly within the species with spiral spurs, the exact mechanics of hawkmoth visitation to flowers with one or more full spiral twists needs further study.
Coevolution and its ecological context
The three long-tongued hawkmoths recorded in this study are involved in pollination interactions with a broader range of flowering plants in Kenya. Agrius convolvuli was also observed feeding widely from other specialized and non-specialized flowers including Crinum macowanii Bak., Ammocharis tinneana Kotschy. & Peyr., Pancratium tenuifolium L., Turraea mombassana C.DC, Gynandropsis gynandra (L.) Briq, Conostomium quadrangulare (Rendle) Cuf., Pavetta abbysinica Fres., and Lippia javanica (Burn.f.) Spreng. Coelonia fulvinotata, the primarily forest-woodland hawkmoth species, was also observed pollinating Clerondendrum rotundifolium Oliv. (Verbenaceae) and Bonatea steudneri (Reichb.f.) Dur. & Schinz, (Orchidaceae), two highly specialized hawkmoth-pollinated species (D. J. Martins, unpublished data). These mutualisms are thus highly asymmetrical, and the hawkmoths can be considered keystone pollinator species (cf. Mills et al., 1993
). Understanding the symmetry of these systems is important for effective conservation of these orchids and other similarly specialized plant species (cf. Ashworth et al., 2004
).
Long-tongued hawkmoths and the long-tubed flowers they pollinate are likely to have coevolved in a diffuse sense (Nilsson et al., 1987
). The asymmetry of the long-tongued hawkmoth pollination system means that the evolution of floral traits of indvidual plants is far more likely to be influenced by particular hawkmoths than is the reverse. Indeed it is questionable whether the orchids represented in this study are numerous enough to have a significant influence on the evolution of their pollinators. On the other hand, the influence of the hawkmoths on the evolution of the orchids has undoubtedly been profound and may have driven speciation in this particular group of orchids. Floral spurs are considered one of the key innovations leading to both increasing specialization in plant–pollinator systems and higher rates of speciation (Hodges, 1995
; Hodges and Arnold, 1995
).
There has been considerable debate about the evolution of long tongues in hawkmoths (Nilsson, 1998
; Jermy, 1999
). Based on studies in flight cages, "swing hovering" was suggested as a predator-evasive behavior of hawkmoths at flowers (Wasserthal, 1997
). It should be noted that no instances of swing hovering were observed during this study during the extensive observations made of hawkmoths foraging in nature on both specialized sphingohilous flowers and generalized flowers, indicating that this is not a typical behavior of long-tongued hawkmoths in eastern Africa.
The hawkmoths observed in this study are highly polyphagous as adults, and under some circumstances individuals will visit the flowers of many different plants in a community, including short-tubed flowers not adapted in any way to hawkmoths (D. J. Martins, unpublished data). Similar observations have been made in Costa Rican dry forest habitats (Haber and Frankie, 1989
; Agosta and Janzen, 2005
). How, then, could such a highly specialized system as that found in the study species be perpetuated, if hawkmoths routinely feed from other flowers? Although hawkmoths are polyphagous, plants such as the orchids in this study establish specialized interactions with them by providing much greater rewards than do other generalist plants in the same community (Table 4; D. J. Martins, unpublished data). Large nectar rewards have also been recorded in other long-tubed hawkmoth-pollinated plants in Kenya, such as Crinum spp., Ammocharis tinneana, Pancratium tenuifolium (Amaryllidaceae), Conostomium quadrangulare (Rubiaceae), and Clerodendrum rotundifolium (Verbenaceae). That the orchids in this study offer relatively large nectar rewards is also evident from the behavior of the hawkmoths, which spend 2–5 s per orchid flower, while visits to generalist flowers rarely last more than 1 s (D. J. Martins, unpublished data).
Conclusion
This study demonstrates that aerangoid orchids in eastern Africa are exclusively pollinated by long-tongued hawkmoths. No more than 1–2 hawkmoth species pollinate a much larger number of specialized orchids. The diversity of the aerangoid orchids (>70 spp. of Aerangis) and their primarily eastern African/Madagascan distribution is suggestive of an adaptive radiation (Simpson, 1953
; Schluter, 2000
) in that both phenotypic divergence and speciation are observable. Though not explored directly in this study, this adaptive radiation may represent an example of plants that have diversified in their pollination strategies.
The pollination success for the orchids of this system is based on potentially diffusively coevolved traits in aerangoid orchids and long-tongued hawkmoths. Spur length and morphology are the driving factors in this particular example, and spurs have been assessed in other taxa to be a "key innovation" (Hodges, 1995
). Notably, the long-spurred aerangoid orchids in eastern Africa are pollinated by a small guild of long-tongued hawkmoths. The mutualism is highly asymmetrical: no more than 1–2 hawkmoth species pollinate a much larger number of specialized orchids. Guilds of plants with deep tubular flowers specialized for pollination by long-proboscid insects, such as hawkmoths and long-proboscid flies, appear to be an important characteristic of the African-Malagasy flora (Nilsson et al., 1998
; Goldblatt and Manning, 2000
). The nectar gradients discovered in the spurs of the orchids in this study may represent a novel reward mechanism that promotes pollinator efficacy in terms of removal and transfer of pollinaria, but further tests of its functional significance still need to be carried out. In general, aerangoid orchids represent an ideal group for further investigation of coevolution and the evolution of specialized plant–pollinator interactions.
FOOTNOTES
1 The authors wish to thank the director and staff of the Mpala Research Centre for logistical support and assistance; the Kenya Orchid Society for help with locating orchid populations and loaning plants for experiments and measurements; the director and staff of the National Museums of Kenya and East African Herbarium for permission to work in their collections;A. Powys, I. and A. Robertson, H. and R. Campbell, C. Ngarachu, P. Masinde, A. Kontos, and S. Miller for providing help in various ways; Nithaya Chetty of the School of Physics and Chemistry for explaining the physics that could account for the nectar concentration gradients; and Rob Raguso and two anonymous reviewers for making very helpful comments that improved the manuscript. This research was supported by a fellowship to Dino J. Martins from the Smithsonian Institution Women's Committee and the Mpala Research Centre and by the University of KwaZulu Natal's School of Biological and Conservation Sciences. 
2 Author for correspondence (e-mail: dinojmv{at}oeb.harvard.edu
; current address: Museum of Comparative Zoology, Harvard University, 26 Oxford Street, Cambridge, MA 02138 USA
LITERATURE CITED
Agnew A. D. Q. Agnew S.. 1994. Upland Kenya wild flowers: a flora of the ferns and herbaceous flowering plants of upland Kenya. East Africa Natural History Society. Nairobi, Kenya..
Agnew J. D.. 1986. Self-compatibility/incompatibility in some orchids of the subfamily Vandoideae. Plant Breeding 97: 183..[CrossRef][Web of Science]
Agosta S. J. Janzen D. H.. 2005. Body size distributions of large Costa Rican dry forest moths and the underlying relationship between plant and pollinator morphology. Oikos 108: 183-193..[CrossRef][Web of Science]
Alexandersson R. Johnson S. D.. 2002. Pollinator-mediated selection on flower-tube length in a hawkmoth-pollinated Gladiolus (Iridaceae). Proceedings of the Royal Society of London, B, Biological Sciences 269: 631-636..
Ashworth L. Aguilar R. Galetto L. Aizen M. A.. 2004. Why do pollination generalist and specialist plant species show similar reproductive susceptibility to habitat fragmentation?. Journal of Ecology 92: 717-719..[CrossRef][Web of Science]
Blundell M.. 1987. Wildflowers of East Africa. William Collins Sons, London, UK..
Carcasson R. H.. 1968. The Sphingidae (hawk moths) of eastern Africa. Ph.D. thesis, University of East Africa, Nairobi, Africa..
Carcasson R. H.. 1976. Revised catalogue of the African Sphingidae (Lepidoptera) with descriptions of the East African species, 2nd ed. E. W. Classey, Faringdon, Oxen, UK..
Cribb P.. 1984. Flora of tropical East Africa: Orchidaceae 271. Royal Botanic Gardens, Kew, UK..
Darwin C.. 1877. The various contrivances by which orchids are fertilized by insects, 2nd ed. John Murray, London, UK..
Dressler R. L.. 1981. The orchids: natural history and classification. Harvard University Press, Cambridge, Massachusetts, USA..
Goldblatt P. Manning J. C.. 2000. The long-proboscid fly pollination system in southern Africa. Annals of the Missouri Botanical Garden 138: 146-170..
Goyret J. Raguso R. A.. 2006. The role of mechanosensory input in flower handling efficiency and learning by Manduca sexta. Journal of Experimental Biology 209: 1585-1593..
Grant V.. 1985. Additional observations on temperate North American hawkmoth flowers. Botanical Gazette 146: 517-520..
Grant V.. 1992. Floral isolation between ornithophilous and sphingophilous species of Ipomopsis and Aquilegia. Proceedings of the National Academy of Sciences, USA 89: 11828-11831..
Haber W. A. Frankie G. W.. 1989. A tropical hawkmoth community: Costa Rican dry forest Sphingidae. Biotropica 21: 155-172..[CrossRef][Web of Science]
Hodges S. A.. 1995. The influence of nectar production on hawkmoth behavior, self pollination, and seed production in Mirabilis multiflora (Nyctaginaceae). American Journal of Botany 82: 197-204..[CrossRef][Web of Science]
Hodges S. A. Arnold M. L.. 1995. Spurring plant diversification: are floral nectar spurs a key innovation?. Proceedings of the Royal Society of London, B, Biological Sciences 262: 113-120..
Janzen D. H.. 1980. When is it coevolution?. Evolution. 34: 611-612..[CrossRef][Web of Science]
Jermy T.. 1999. Deep flowers for long tongues: a final word. Trends in Ecology and Evolution 14: 34-34..[CrossRef]
Johnson S. D.. 1995. Observations of hawkmoth pollination in the South African orchid Disa cooperi. Nordic Journal of Botany 15: 121-125..[Web of Science]
Johnson S. D. Edwards T. J. Carbutt C. Potgieter C.. 2002. Specialization for hawkmoth and long-proboscid fly pollination in Zaluzianskya section Nycterinia (Scrophulariaceae). Botanical Journal of the Linnean Society 138: 17-27..[CrossRef][Web of Science]
Johnson S. D. Liltveld W. R.. 1997. Hawkmoth pollination of Bonatea speciosa (Orchidaceae) in a South African coastal forest. Nordic Journal of Botany 17: 5-10..[Web of Science]
Johnson S. D. Neal P. R. Peter C. I. Edwards T. J.. 2004. Fruiting failure and limited recruitment in remnant populations of the hawkmoth-pollinated tree Oxyanthus pyriformis subsp. pyriformis (Rubiaceae). Biological Conservation 120: 31-39..[CrossRef][Web of Science]
Johnson S. D. Nilsson L. A.. 1999. Pollen carryover, geitonogamy, and the evolution of deceptive pollination systems in orchids. Ecology 80: 2607-2619..[CrossRef][Web of Science]
Kaiser R.. 1993. The scent of orchids: olfactory and chemical investigations. Elsevier, Amsterdam, Netherlands..
Kelber A.. 1996. Colour learning in the hawkmoth Macroglossum stellatarum. Journal of Experimental Biology 199: 1127-1131..[Abstract]
Kelber A.. 1997. Innate preferences for flower features in the hawkmoth Macroglossum stellatarum. Journal of Experimental Biology 200: 827-836..[Abstract]
Kelber A. Balkenius A. Warrant E. J.. 2002. Scotopic colour vision in nocturnal hawkmoths. Nature 419: 922-925..[CrossRef][Medline]
Kelber A. Pfaff M.. 1997. Spontaneous and learned preferences for visual flower features in a diurnal hawkmoth. Israel Journal of Plant Sciences 45: 235-245..[Web of Science]
Kingston A. J. Nummelin M.. 1998. Seasonality and abundance of Sphingids in a garden on the lower slopes of the Uluguru Mountains in Morogoro Township in Tanzania. Journal of East African Natural History 87: 213-220..[CrossRef]
Koopowitz H. Marchant T. A.. 1998. Postpollination nectar reabsorption in the African epiphyte Aerangis verdickii (Orchidaceae). American Journal of Botany 85: 508-512..[Abstract]
Luyt R. Johnson S. D.. 2002. Postpollination nectar reabsorption and its implications for fruit quality in an epiphytic orchid. Biotropica 34: 442-446..[Web of Science]
McNeely C. Singer M. C.. 2001. Contrasting the roles of learning in butterflies foraging for nectar and oviposition sites. Animal Behaviour 61: 1-7..[CrossRef][Web of Science][Medline]
Miller R. B.. 1981. Hawkmoths and the geographic patterns of floral variation in Aquilegia caerulea. Evolution 35: 763-774..[CrossRef][Web of Science]
Miller R. B.. 1985. Hawkmoth pollination of Aquilegia-Chrysantha Ranunculaceae in southern Arizona USA. Southwestern Naturalist 30: 69-76..[CrossRef]
Miller W. E.. 1997. Diversity and evolution of tongue length in hawkmoths (Sphingidae). Journal of the Lepidopterists' Society 51: 9-31..
Mills L. S. Soulé M. E. Doak D. F.. 1993. The keystone-species concept in ecology and conservation. BioScience 43: 219-224..[CrossRef][Web of Science]
Ng'weno F.. 1981. Hawkmoths in my garden. Bulletin of the East Africa Natural History Society. September–October, 1–1..
Nilsson L. A.. 1988. The evolution of flowers with deep corolla tubes. Nature 334: 147-149..[CrossRef]
Nilsson L. A.. 1992. Orchid pollination biology. Trends in Ecology and Evolution 7: 255-259..[CrossRef]
Nilsson L. A.. 1998. Deep flowers for long tongues. Trends in Ecology and Evolution 13: 259-260..[CrossRef]
Nilsson L. A. Jonsson L. Ralison L. Randrianjohany E.. 1987. Angraecoid orchids and hawkmoths in central Madagascar: specialized pollination systems and generalist foragers. Biotropica 19: 310-318..[CrossRef][Web of Science]
Nilsson L. A. Rabakonandrianina E. Pettersson B.. 1992. Exact tracking of pollen transfer and mating in plants. Nature 360: 666-667..[CrossRef]
Owen D. F.. 1969. Species diversity and seasonal abundance in tropical Sphingidae (Lepidoptera). Proceedings of the Royal Entomological Society of London, A 44: 162-168..
Pinhey E.. 1962. Hawk moths of central and southern Africa. National Museums of Southern Rhodesia, Longmans, South Africa..
Robsertson I. A. D.. 1977. Records of insects taken at light traps in Tanzania. V.–seasonal changes in catches and effect of the lunar cycle on hawkmoths of subfamily Asemanophorinae (Lepidoptera: Sphingidae). Miscellaneous Report No. 31. Centre for Overseas Pest Research, College House, London, UK..
Schluter D.. 2000. The ecology of adaptive radiation. Oxford University Press, Oxford, UK..
Simpson G. G.. 1953. The major features of evolution. Columbia University Press, New York, New York, USA..
Stewart J.. 1996. Orchids of Kenya. Timber Press, Portland, Oregon, USA..
Tremblay R. L. Ackerman J. D. Zimmerman J. K. Calvo R. N.. 2005. Variation in sexual reproduction in orchids and its evolutionary consequences: a spasmodic journey to diversification. Biological Journal of the Linnean Society 84: 1-54..[CrossRef][Web of Science]
Wasserthal L. T.. 1997. The pollinators of the Malagasy star orchids Angraecum sesquipedale, A. sororium and A. compactum and the evolution of extremely long spurs by pollinator shift. Botanica Acta 110: 343-359..[Web of Science]
Weiss M. R. Papaj D. R.. 2003. Colour learning in two behavioural contexts: how much can a butterfly keep in mind?. Animal Behaviour 65: 425-434..[CrossRef][Web of Science]
CiteULike Complore Connotea Del.icio.us Digg Facebook Reddit Technorati Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Micheneau, J. Fournel, B. H. Warren, S. Hugel, A. Gauvin-Bialecki, T. Pailler, D. Strasberg, and M. W. Chase Orthoptera, a new order of pollinator Ann. Bot., January 11, 2010; (2010) mcp299v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. O. Schlumpberger, A. A. Cocucci, M. More, A. N. Sersic, and R. A. Raguso Extreme variation in floral characters and its consequences for pollinator attraction among populations of an Andean cactus Ann. Bot., June 1, 2009; 103(9): 1489 - 1500. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
