Oldweb.northampton.ac.uk

FORUM is intended for new ideas or new ways of interpreting existing information. It provides a chance for suggesting hypotheses and for challenging current thinking onecological issues. A lighter prose, designed to attract readers, will be permitted. Formal research reports, albeit short, will not be accepted, and all contributions should be concise with a relatively short list of references. A summary is not required.
Latitudinal trends in plant-pollinator interactions: are tropical
plants more specialised?
Jeff Ollerton and Louise Cranmer, School of En7ironmental Science, Uni7. College Northampton, Park Campus,
Northampton, NN2 7AL, UK ( jeff.ollerton@northampton.ac.uk).
The increase in richness of species and higher taxa going from
diversity and consequent finer division of resources, polar higher to lower latitudes is one of the most studied global
areas because of low species diversity and therefore a lack biogeographical patterns. Latitudinal trends in the interactions
of opportunity for species to be more generalised. Are between species have, in contrast, hardly been studied at all,
probably because recording interactions is much less straightfor-
there any data for global patterns of species interactions ward than counting species. We have assembled two independent
data sets which suggest that plant-pollinator interactions are not
Few quantitative studies have explicitly addressed the more ecologically specialised in the tropics compared to temper-
question of how the specificity of species interactions ate latitudes. This is in contrast to a prevailing view that tropical
ecological interactions tend towards higher specificity than their
varies with latitude. For example, Scriber (1973) and temperate counterparts.
Price (1980) looked at specialisation in larval feeding inLepidoptera, Beaver (1978) dealt with bark and ambrosiabeetles, Hawkins (1990) and Porter and Hawkins (1998) Latitudinal trends in biodiversity are well known for studied global patterns of parasitoid numbers per insect many groups of organisms, with taxon diversity being host, whilst Rohde (1978) focused on latitudinal trends in fish parasites. Their findings will be considered later, (MacArthur 1972, Rohde 1992, Gaston and Williams but the question of tropical ecological specialisation 1996). In contrast, we know much less about latitudinal remains largely unresolved for most categories of inter- trends in the biodiversity of species interactions. For action and the functional groups involved in these example, as one moves from temperate to tropical latitudes, do predator-prey, parasite-host or mutualistic Despite their importance in most terrestrial ecosystems (Kearns and Inouye 1997), we possess an inadequate (defined as the number of species involved in the interac- knowledge of the broad biogeographic patterns of plant- tion, as distinct from morphological or evolutionary pollinator interactions and the underlying causes of any specialisation; Waser et al. 1996, Armbruster et al. 2000)? pattern (Johnson and Steiner 2000). An initial reading of The proposal that resources are divided more finely the literature would suggest that there is a consensus amongst a greater number of species in the tropics, amongst pollination biologists that tropical pollination compared to temperate communities (MacArthur 1972, systems are more ecologically specialised than temperate Janzen 1973), suggests that tropical organisms should systems (Johnson and Steiner 2000) but there are almost indeed be more ecologically specialised. However, low no data to support this assertion, and only limited data species diversity in very high latitude areas may also lead to refute it (Kevan and Baker 1983).
to apparent ecological specialisation in species interac-tions. In this case, the resulting latitudinal trend wouldbe hump-backed – high specificity of interactions in the The data sets
tropics and towards polar regions, with much lowerspecificity (greater generalisation) at temperate latitudes.
As a step towards understanding whether pollination Interestingly, the extremes of the gradient would show systems show a significant latitudinal trend in speciali- greater specialisation in interactions for diametrically sation, we have assembled two independent data sets at opposite reasons – the tropics because of high species different taxonomic/ecological scales, full details of which are given in Appendices 1 and 2. The first data variance in this regression model is explained by differ- set is at the scale of the plant community and comprises ences in sampling effort between communities. Clearly 27 published and unpublished surveys of plant-flower tropical community pollination studies suffer from un- visitor interactions in 35 communities at different lati- der sampling of the true diversity of flower visitors per tudes. From these studies we extracted information on plant species (though note that latitude and sampling the latitude at which the study was undertaken (deci- effort are not directly correlated – Pearson’s Product malised for the purposes of analysis), mean number of Moment Correlation: r = 0.05, df = 33, p = 0.78).
species of flower visitors per plant species (most of these The distribution of the asclepiad data set is highly studies recorded flower visitors rather than pollinators non-normal and untransformable and therefore violates per se; however, number of flower visitors is strongly the requirements of multiple regression analysis. To correlated with number of pollinators and this should take account of sampling effort for these data we have therefore be an appropriate proxy [Ollerton, unpubl.]), corrected number of pollinators per plant species by number of plant species studied and sampling effort dividing by number of days sampling for the subset of (number of field days of observation). The latter vari- data where this is known (Fig. 1c). Correcting for able was in some studies explicitly stated and in others sampling effort in this way removes any correlation was estimated from the published information.
between latitude and pollinator specialisation. Once The second data set consists of 103 published and again, the apparently more specialised tropical species unpublished studies of pollinators of species of asclepi- suffer from under sampling of pollinators.
ads (subfamily Asclepiadoideae of the Apocynaceae Two completely independent data sets, at two differ- sensu Endress and Bruyns 2000). This is part of the ent taxonomic scales, show precisely the same result, that tropical plants are, on average, no more ecologi- bayreuth.de/departments/planta2/wgl/fsigrid2.html). As cally specialised in their pollination systems than tem- in the first data set, we extracted information on lati- perate species. We conclude that the apparent trend tude, number of pollinators per plant species (in this towards more specialised pollination systems for tropi- case, a much more straightforward variable as asclepi- cal plants shown in Fig. 1a and b is an artefact of ads possess aggregations of pollen (pollinia) that me- sampling bias and that there is no significant latitudinal trend in the specificity of plant-pollinator relationships.
identification of pollinators much easier – see Ollerton How do our results compare to the previously pub- and Liede 1997) and number of days of observation, lished studies cited earlier. In particular, is there any which was available for only 59 of the 103 studies.
evidence from other work that the humpbacked latitu-dinal trend may occur in some interactions? Thesestudies have looked at a range of organisms and typesof interaction and have uncovered a variety of relation- Results and discussion
ships between latitude and ecological specificity. Scriber(1973) was probably the first worker to confront quan- Initial analyses of these data sets suggests that pollina- titatively the problem of temperate versus tropical spe- tion systems do indeed become more specialised moving cialisation, in a study of larval host plant use in from temperate latitudes towards the tropics (Fig. 1a Papilionidae (Lepidoptera). His analysis showed that a and b). In both the community and asclepiad data sets higher proportion of temperate species could be consid- there is a significant positive relationship between lati- ered generalist compared to tropical species. Scriber’s tude and number of pollinators/flower visitors per plant definition of generalist taxa was ‘‘…those species feed- species. This is also true if the data are separated into ing on more than one taxonomic family of plants…’’.
northern and southern hemispheres (data not pre- This may be considered a rather broad definition of sented). However, closer analysis reveals that this pat- ‘‘generalised’’ and, intriguingly, Price (1980) presented tern is misleading. The various studies included within data that suggested that tropical butterflies tended to be the community and asclepiad data sets varied consider- no more host specific than temperate species. Rohde ably in the sampling effort undertaken to observe and (1978) found that tropical taxa of marine platyhelminth record flower visitors. To take account of this we have fish parasites in the group Digenea were more host used sampling effort per plant species together with specific than temperate taxa, but that this was not so in latitude as independent variables in a multiple regres- the Monogenea. Beaver (1978) showed that bark and sion analysis of the community data set (Table 1).
ambrosia beetles (Coleoptera: Scolytidae and Platypo- Forty percent of the variation in mean number of didae) are actually less host specific in the tropics flower visitors per plant species is explained by this compared to temperate communities, a pattern that he stepwise multiple regression model. However, only 4% considered may be explained by the low population of this variation results from the latitude at which the densities of host trees in the tropics. Hawkins (1990) study was conducted (and this is only significant at studied parasitoids of phytophagous insects with differ- p = 0.075 for the t-ratio test). The remaining 36% of the ent feeding ecologies and showed that those parasitising exposed hosts tended to be more host specific in the same’’. This conclusion is confirmed by the data that we tropics, whilst no such pattern was apparent for para- have presented in this paper. Tropical communities sitoids utilising hosts concealed in plant tissue. Clearly, provide some of the best examples of close co-evolved different categories of species interaction and different plant-pollinator relationships and in absolute terms do groups of taxa may or may not show increased special- contain a higher number of plants with specialised pollination systems. However, tropical plant assem- A literature review by Kevan and Baker (1983) con- blages are on average many times more species-rich than cluded that ‘‘…from the arctic and alpine areas to the their temperate counterparts and so may not in fact lowland tropics, it appears that the frequency of occur- possess disproportionately more ecologically specialised rence of specialised pollination syndromes is about the pollination systems than temperate assemblages.
Fig. 1. Relationshipsbetween latitude andpollinator specialisationfor the communitysurvey and asclepiaddata sets.
a. Community surveysof plant-flower visitorrelationships. Meannumber of species offlower visitors per plantspecies has been logtransformed. Pearson’sproduct momentcorrelation: r = 0.33,df = 33, p = 0.051.
b. Pollinators ofasclepiads. Spearmanrank correlation:r = 0.33, n = 91,p = 0.002. c. Pollinatorsof asclepiads, correctedfor sampling effort.
Spearman rankcorrelation: r = 0.09,n = 59, p = 0.51.
Table 1. Results of stepwise multiple regression on mean number of species of flower visitors per plant species for thecommunity data set. All variables were natural log transformed.
Problems with the data sets
plant species, localities and years and so there is nosimple ‘‘rule of thumb’’ which would allow us to apply The type of analysis that we have presented, in which a simple correction. We have therefore opted to use a largely pre-existing data are evaluated in relation to a range of days of sampling effort to test how a saturat- question which they were not primarily collected to ing sampling function would affect our conclusions. We address, can be fraught with statistical problems. We repeated the analysis of the asclepiad data set using have identified two possible causes for concern within sampling saturation points between 1 day and 60 days the two data sets, which we detail below.
of sampling effort (Table 2). This covered the range of The first statistical problem concerns the phyloge- numbers of days of actual sampling effort undertaken netic relatedness of the plants and pollinators in the by the various studies in Appendix 2. The analysis analysis. It is acknowledged (and debated) that possible involved repeating the Spearman rank correlations be- phylogenetic biases must be taken into consideration in tween number of species of pollinator (corrected for any comparative analysis (Harvey and Pagel 1991).
sampling effort) and latitude and successively restricting However, the community survey data set spans such a the maximum number of days by which number of wide range of plant and animal genera, orders and pollinators was corrected to 1, 2 … 10 … 20 … up to classes that a formal phylogenetically-corrected regres-sion is not possible. Whether it is required for such a 60 days. Low levels of maximum sampling effort (less phylogenetically broad spread of taxa is arguable. In than 10 days) yielded results not quantitatively different relation to the asclepiad data set, a robust molecular from that shown in Fig. 1b, with statistically significant generic-level phylogeny of the group is not yet avail- relationships between latitude and number of species of able. Therefore, whilst we recognise that the phyloge- pollinator. That is to say, correcting by a maximum of netic architecture of this data set may be a statistical only a modest sampling effort is approximately similar problem (for example, the higher latitude data mainly to not correcting the data at all, a not unexpected come from North American Asclepias species) we can- result. The statistically significant correlation disap- not at the present time allow for this.
pears when using more realistic saturation levels of The second statistical problem specifically concerns the asclepiad data set. In order to correct for different Table 2. Spearman rank correlations of latitude versus num-ber of species of pollinators per plant species corrected by sampling efforts across studies, for each plant the num- sampling effort for a range of sampling effort saturation ber of recorded pollinators was divided by the number points. N = 59 in all cases, except the uncorrected analysis, of days of sampling. This correction assumes a linear relationship between sampling effort and number of pollinators per plant species. In reality the relationshipis likely to be saturating, with records of new pollina- tors declining to zero at some point during the observa- tion period. If the relationship between sampling effort and number of observed pollinators is indeed saturat- ing, our simple correction would result in an under estimate of the number of pollinators per plant species expected from a given level of sampling effort. It is impossible to say what the exact sampling saturation point is as this information is never presented in studies of plant-pollinator interactions. In a recent survey of asclepiad pollinators at a site in South Africa, we had sampled all of the pollinators of some species in as little as 10 days, though for other species we were still recording new pollinators after 30 days (Ollerton et al.
in prep.). Sampling saturation points (beyond which no new pollinators are recorded) are likely to vary between exploring latitudinal trends in plant-pollinator interac-tions, these data sets are as good as any that couldbe currently assembled. We hope that by publishingthis study we will stimulate interest in the question oftropical versus temperate specialisation in ecologicalinteractions and that future researchers will obtaingrants large enough to allow dedicated data collectionthat will tackle this question. Until such time, thesedata sets must suffice.
Acknowledgements – The ideas presented in this paper havebenefited from discussion with many colleagues. We wouldparticularly like to thank Scott Armbruster, Kevin Gaston,David Inouye, Steve Johnson, Duncan McCollin, Jane Mem-mott, Paul Neal, Jens Olesen and Nick Waser and an anony-mous reviewer. We also thank Steve Johnson, Sigrid Liede,Jane Memmott, Jens Olesen, Anton Pauw and Milene Vieirafor providing us with unpublished data. We are grateful to thefollowing organisations for providing funding which con-tributed to some the results in the paper: The Royal Society,The Leverhulme Trust, Church and Co. PLC, The BiodiversityTrust, The Percy Sladen Memorial Fund and The RoyalEntomological Society.
Fig. 2. The relationship between number of days sampling References
effort and the number of flower visitors/pollinators per plantspecies in (a) the community data set; and (b) the asclepiad Armbruster, W. S., Fenster, C. B. and Dudash, M. R. 2000.
Pollination ‘‘principles’’ revisited: specialization, pollina-tion syndromes and the evolution of flowers. – Det NorskeVidenskaps-Akademi I. Mat-Naturv. Klasse Skrifter Ny.
greater than 10 days sampling (Table 2) and confirm Beaver, R. A. 1978. Host specificity of temperate and tropical In conclusion, correcting the data for sampling ef- animals. – Nature 281: 1139 – 1141.
Endress, M. E. and Bruyns, P. V. 2000. A revised classification fort using a realistic saturating function (by which we of the Apocynaceae s.l. – Bot. Rev. 66: 1 – 56.
consider that 10 days or less of observation is un- Fishbein, M. and Venable, D. L. 1996. Diversity and temporal likely to identify all of the pollinators of even a mod- change in the effective pollinators of Asclepias tuberosa. –Ecology 77: 1061 – 1073.
erately generalised species) does not affect the results Gaston, K. J. and Williams, P. H. 1996. Spatial patterns in obtained when a linear, non-saturating correction is taxonomic diversity. – In: Gaston, K. J. (ed.), Biodiversity: applied. This raises quite a fundamental issue in rela- a biology of numbers and difference. Blackwell Scientific,pp. 202 – 229.
tion to studying pollination ecology – when can we Hawkins, B. A. 1990. Global patterns of parasitoid assemblage be sure that we have identified all of the pollinators size. – J. Anim. Ecol. 59: 57 – 72.
of a plant? The annual fluctuations in pollinator Harvey, P. H. and Pagel, M. 1991. The comparative method in abundances that are a feature of many plant-pollina- evolutionary biology. – Oxford Univ. Press.
Janzen, D. H. 1973. Comments on host-specificity of tropical tor systems (see, amongst many potential examples, herbivores and its relevance to species richness. – In: Pettersson 1991, Fishbein and Venable 1996, Lam- Heywood, V. H. (ed.), Taxonomy and ecology. Academic born and Ollerton 2000) suggests that a time scale of Johnson, S. D. and Steiner, K. E. 2000. Generalization versus years to decades may be necessary before a complete specialization in plant pollination systems. – Trends Ecol.
list of pollinators is obtained for generalist pollination systems. This is reinforced by a crude analysis com- Kearns, C. A. and Inouye, D. W. 1997. Pollinators, flowering plants, and conservation biology. – BioScience 47: 297 – paring sampling effort to number of identified polli- nators in the community and asclepiad data sets Kevan, P. G. and Baker, H. G. 1983. Insects as flower visitors presented here (Fig. 2a and b). In both of the data and pollinators. – Ann. Rev. Ent. 28: 407 – 453.
Lamborn, E. and Ollerton, J. 2000. Experimental assessment sets there is no suggestion of a levelling off of num- of the functional morphology of inflorescences of Daucus bers of identified pollinators as sampling effort in- carota (Apiaceae): testing the ‘‘fly catcher effect’’. – Funct.
We have attempted to be honest about the limita- MacArthur, R. H. 1972. Geographical ecology. – Harper and tions of our data sets and would argue that apprecia- Ollerton, J. and Liede, S. 1997. Pollination systems in the tion of these problems does not negate their value, Asclepiadaceae: a survey and preliminary analysis. – Biol.
nor the value of our analyses. For the purposes of Pettersson, M. W. 1991. Pollination by a guild of fluctuating marine Monogenea and Digenea. – Mar. Biol. 47: 125 – 134.
moth populations: option for unspecialization in Silene Rohde, K. 1992. Latitudinal gradients in species diversity: the 6ulgaris. – J. Ecol. 79: 591–604.
search for the primary cause. – Oikos 65: 514 – 527.
Porter, E. E. and Hawkins, B. A. 1998. Patterns of diversity for Scriber, J. M. 1973. Latitudinal gradients in larval feeding aphidiine (Hymenoptera: Braconidae) parasitoid assem- specialization of the world Papilionidae (Lepidoptera). – blages on aphids (Homoptera). – Oecologia 116: 234 – 242.
Price, P. W. 1980. Evolutionary biology of parasites. – Prince- Waser, N. M., Chittka, L., Price, M. V. et al. 1996. Generaliza- tion in pollination systems, and why it matters. – Ecology Rohde, K. 1978. Latitudinal differences in host-specificity of Appendix 1. Studies included in the community-level data set.
Appendix 2. Studies used in the asclepiad data set. Note that data for sampling effort are not available for all studies.
Ceropegia lushi var. acuminata Oxypetalum alpinum var. alpinum Oxypetalum banksii subsp. banksii Ali, T. 1994. Pollination ecology of some asclepiads (Asclepiadaceae) from Pakistan. – Unpublished Ph.D. thesis, Univ. of Karachi.
Arroyo, M. T. K., Primack, R., Armesto, J. 1982. Community studies in pollination ecology in the high temperate Andes of Chile. I. Pollination mechanisms and altitudinal variation.
– Am. J. Bot. 69: 82 – 97.
Barrett, S. C. H. and Helenurm, K. 1987. The reproductive biology of boreal forest herbs, I. Breeding systems and pollination. – Can. J. Bot. 65: 2036 – 2046.
Bhatnagar, S. 1986. On insect adaptations for pollination in some asclepiads of Central India. – In: Kapil, R. P. (ed.) Pollination biology – an analysis. Inter-India Publications,New Delhi, pp. 37 – 57.
Betz, R. F., Struven, R. D., Wall, J. E. and Heitler, F. B. 1994. Insect pollinators of 12 milkweed (Asclepias) species. – In: Wickett, R. G., Lewis, P. D., Woodliffe, A. and Pratt,P. (eds.) Proc. Thirteenth N. Am. Prairie Conf. Dept of Parks and Recreation, Ontario, Canada, pp. 45 – 60.
Bosch, J., Retana, J. and Cerda´, X. 1997. Flowering phenology, floral traits and pollinator composition in a herbaceous Mediterranean plant community. – Oecologia 109: 583 – 591.
Burkill, I. H. 1897. Fertilization of some spring flowers on the Yorkshire coast. – J. Bot. 35: 92 – 189.
Chaplin, S. J. and Walker, J. L. 1982. Energetic constraints and adaptive significance of the floral display of a forest milkweed. – Ecology 63: 1857 – 1870.
Chaturvedi, S. K. and Pant, D. D. 1986. Further studies in the pollination of some Indian asclepiads. – Bull. Bot. Survey India 28: 23 – 30.
Clements, F. E. and Long, F. L. 1923. Experimental pollination – an outline of the ecology of flowers and insects. – Carnegie Institution.
Drapalik, D. J. 1969. A biosystematic study of the genus Matelea in the southeastern United States. – Unpublished Ph.D. thesis, Univ of North Carolina.
Eberling, H. and Olesen, J. M. 1999. The structure of a high latitude plant-flower visitor system: the dominance of flies. – Ecography 22: 314 – 323.
Fishbein, M. and Venable, D. L. 1996. Diversity and temporal change in the effective pollinators of Asclepias tuberosa. – Ecology 77: 1061 – 1073.
Forster, P. I. 1989. Pollination of Marsdenia fraseri (Asclepiadaceae) by Metriorrhynchus lateralis (Coleoptera: Lycidae). – Coleopt. Bull. 43: 311 – 312.
Forster, P. I. 1991. A possible identification for ‘‘Pollinia attached to adult anopheline mosquitoes from northern Australia’’. – Entom. Soc. Queensl. News Bull. 18: 113.
Forster, P. I. 1992. Insects associated with the flowers of Marsdenia cymulosa Benth. (Asclepiadaceae) and their possible role in pollination. – Aust. Entom. Mag. 19: 45 – 47.
Hagerup, O. 1932. On pollination in the extremely hot air at Timbuctoo. – Dansk. Bot. Ark.8: 1 – 20.
Herrera, J. 1988. Pollination relationships in southern Spanish mediterranean shrublands. – J. Ecol. 76: 274 – 287.
Hocking, B. 1968. Insect-flower associations in the high Arctic with special reference to nectar. – Oikos 19: 359 – 388.
Inoue, T., Kato, M. and Kakutani, T. 1990. Insect-flower relationship in the temperate deciduous forest of Kibune, Kyoto: an overview of the flowering phenology and seasonalpattern of insect visits. – Contr. Biol. Lab. Kyoto Univ. 27: 377 – 463.
Inouye, D. W. and Pike, G. H. 1988. Pollination biology in the snowy mountains of Australia: comparisons with montane Colorado, USA. – Aust. J. Ecol. 13: 191 – 210.
Jonkers, B. 1990. Carallumas – gems of the mountainside. – Petr. Develop. Oman News 2: 7 – 11.
Jonkers, B. 1993. De bestuivers van succulenten. – Succulenta 72: 268 – 275.
Kanstrup, J. and Olesen, J.M. 2000. Plant-flower visitor interactions in a neotropical rain forest canopy: community structure and generalisation level. – Det NorskeVidenskaps-Akademi I. Matematisk-Naturvitenskapelig Klasse, Avhandlinger, Ny Serie 39: 33 – 41.
Kato, M., Kakutani, T., Inoue, T. and Itino, T. 1990. Insect-flower relationship in the primary beech forest of Ashu, Kyoto: an overview of the flowering phenology and seasonalpattern of insect visits. – Contr. Biol. Lab. Kyoto Univ. 27: 309 – 375.
Kevan, P. G. 1970. High arctic insect-flower relations: the inter-relationships of arthropods and flowers at Lake Hazen, Ellesmere Island, Northwest Territories, Canada. –Unpublished Ph.D. thesis, Univ. of Alberta, Canada.
Kunze, H. and Liede, S. 1991. Observations on pollination in Sarcostemma (Asclepiadaceae). – Plant Syst. Evol. 178: 95 – 105.
Liede, S. 1994. Some observations on pollination in Mexican Asclepiadaceae. – Madron˜o 41: 266 – 276.
Liede, S. and Whitehead, V. 1991. Studies in the pollination biology of Sarcostemma 6iminale R.BR. sensu lato. – South Afr. J. Bot. 57: 115–122.
Lynch, S. P. 1977. The floral ecology of Asclepias solanoana woods. – Madrofio 24: 159 – 177.
Macior, L. W. 1965. Insect adaptation and behaviour in Asclepias pollination. – Bull. Torrey Bot. Club 92: 114 – 126.
Memmott, J. 1999. The structure of a plant-pollinator food web. – Ecol. Lett. 2: 276 – 280.
Meve, U. and Liede, S. 1994. Floral biology and pollination in stapeliads – new results and a literature review. – Plant Syst. Evol. 192: 99 – 116.
Moldenke, A. R. and Lincoln, P. G. 1979. Pollination ecology in montane Colorado: a community analysis. – Phytologia 42: 349 – 379.
Momose, K., Yumoto, T., Nagamitsu, T. et al. 1998. Pollination biology in a lowland dipterocarp forest in Sarawak, Malaysia. I. Characteristics of the plant-pollinator communityin a lowland dipterocarp forest. – Am. J. Bot. 85: 1477 – 1501.
Mosquin, T. and Martin, J. E. H. 1967. Observations on the pollination biology of plants on Melville Island, N.W.T., Canada. – Can. Field-Nat. 81: 201 – 205.
Nel, M. 1995. Rare and interesting plants of the Namib desert, part 2: three desert plants. – Veld and Flora 81: 14 – 15.
Pant, D. D., Nautiyal, D. D. and Chaturvedi, S. K. 1982. Pollination ecology of some Indian asclepiads. – Phytomorphology 32: 302 – 313.
Pauw, A. 1998. Pollen transfer on birds’ tongues. – Nature 394: 731 – 732.
Payson, E. 1916. The pollination of Asclepias cryptoceras. – Bot. Gaz. 61: 72 – 74.
Percival, M. 1974. Floral ecology of coastal scrub in southeast Jamaica. – Biotropica 6: 104 – 129.
Petanidou, T. 1991. Pollination ecology in a phryganic ecosystem. – Unpublished Ph.D. thesis, Univ of Thessaloniki.
van der Pijl, L. 1954. Xylocopa and flowers in the tropics I, II and III. – Konink. Nederl. Akad. van Wetens, Proc. C. 57: 413 – 562.
Piper, R. G., Sweeney, A. W. and Gibbons, D. S. 1990. Pollinia attached to adult anopheline mosquitoes from northern Australia. – Entom. Soc. Queensl. News Bull. 18: 83 – 84.
Primack, R. B. 1983. Insect pollination in the New Zealand mountain flora. – New Zeal. J. Bot. 21:317 – 333Ramakrishna, T. M. and Arekal, G. D. 1979. Pollination biology of Calotropis gigantea (L.) R.Br. – Curr. Sci. 48: 212 – 213.
Ramakrishna, T. M. and Arekal, G. D. 1982 – 83. Pollination in Pergularia daemia (Asclepiadaceae). – J. Mysore Univ. Sect. B. 29: 1 – 3.
Rarnirez, N. 1989. Biologia de polinisacion en una comunidad arbustiva tropical de la Alta Guayana Venezolana. – Biotropica 21:319 – 330.
Rarmirez, N. and Brito, Y. 1992. Pollination biology in a palm swamp community in the Venezuelan Central Plains. – Bot. J. Linn. Soc. 110: 277 – 302.
Robertson, C. 1891. Flowers and insects, Asclepiadaceae to Scrophulariaceae. – Trans. St. Louis Acad. Sci. 5: 569 – 598.
Robertson, C. 1928. Flowers and insects: lists of visitors of four hundred and fifty-three flowers. – Privately published, Carlinville, Illinois.
Sabrosky, C. W. 1987. A new species of Leptometopa (Diptera, Milichiidae) from Madagascar pollinating Ceropegia (Asclepiadaceae). – Proc. Ent. Soc. Washington 89: 242 – 243.
Skutch, A. F. 1988. Flowering and seed-production of Fischeria fimebris (Asclepiadaceae). – Brenesia 30: 13 – 17.
Small, E. 1976. Insect pollinators of the Mer Bleue peat bog of Ottawa. – Can. Field-Nat. 90: 22 – 28.
Struck, M. 1995. Land of blooming pebbles: flowers and their pollinators in the Knersvlakte. – Aloe 32: 56 – 64.
Vieira, M. F. and Shepherd, G. F. 1999. Pollinators of Oxypetalum (Asclepiadaceae) in southeastern Brazil. – Rev. Brasil. Biol. 59: 693 – 704.
Wanntorp, H. E. 1974. Calotropis gigantea (Asclepiadaceae) and Xylocopa tenuiscapa (Hymenoptera: Apidae). – Svensk Bot. Tidsk. 68: 25 – 32.
Willis, J. C. and Burkill, I. H. 1903a. Flowers and insects in Great Britain. Part II. – Ann. Bot. 17: 313 – 349.
Willis, J. C. and Burkill, I. H. 1903b. Flowers and insects in Great Britain. Part III. – Ann. Bot. 17: 539 – 570.
Willis, J. C. and Burkill, I. H. 1908. Flowers and insects in Great Britain. Part IV. – Ann. Bot. 22: 603 – 649.
Willson, M. F., Bertin, R. I. and Price, P. W. 1979. Nectar production and flower visitors of Asclepias 6erticillata. – Am. Midl. Nat. 102: 23–35.

Source: http://oldweb.northampton.ac.uk/aps/env/lbrg/journals/papers/OllertonCranmerLatitudinalTrendsOikos2002.pdf