Extrafloral nectaries vs ant-Homoptera mutualisms: a comment on Becerra and Venable.

Becerra and Venable (1989) proposed that the function of extrafloral nectaries (EFNs) may not exclusively be to attract ants which defend plants against herbivores. They questioned the protection hypothesis and proposed that EFNs may also function to defend plants from ant-Homoptera mutualisms. Arguments for their hypothesis were e.g. the lacking universality of ant protection for plants and the negative impact of ant-tended homopterans on their host plants. In general, effects of homopterans on plants are negative but the damages of ant-homopteran associations are difficult to assess. Becerra and Venable cited a few examples of beneficial by-products of these associations in form of protection for the plants by the visiting ants. However, they considered only facultative interactions. In most of the highly developed ant-plant mutualisms ants cultivate homopterans while EFNs are absent. Among plant species of 36 genera, for example, which possess permanent domatia formed by hollow stems, only four have EFNs, but in 18 of them ants were recorded to keep homopterans (Huxley 1986). In the eight best-studied obligate myrmecophytic associations (Table 1) most plant partners offer nesting space but no food in form of nectar. The only two exceptions are Acacia spp. and Leonardoxa africana. In this respect the genus Macaranga (Euphorbiaceae) is especially interesting. In SE Asia it comprises the full range of species which are obligatorily inhabited by ants to those which have only weak and unspecific associations with ants (Fiala et al. 1989, 1990). Most of the facultatively associated Macaranga species possess EFNs on the leaf margins and in addition on the lamina. Obligate myrmecophytes in this genus do not secrete sugar-containing fluids (Fiala and Maschwitz, in press) and colonizing ants keep scale insects inside the hollow stems of their Macaranga host plants. Contrary to Becerra's and Venable's hypothesis in this case a reduction of nectary production appears to be favoured. The hypothesis of superiority of extrafloral nectar to honeydew remains to be tested in facultative associations, but my preliminary observations in Malaysia do not support this view. Becerra and Venable list a number of questions which should be addressed experimentally. I will briefly comment here on the first one: Is it generally true that ants neglect Homoptera when an alternative sugar source is available? During studies on the tree genus Macaranga in Peninsular Malaysia I looked for plants with EFNs and the occurrence of homopterans on these plants. The main study sites were secondary forests and disturbed habitat which were rich in light-demanding, fast growing plants. Six common species which possess EFNs were studied: Macaranga gigantea, M. tanarius, Mallotus paniculatus, M. macrostachys, Endospermum diadenum and Croton argyratus (all Euphorbiaceae). I checked 10 specimens each for occurrence of homopterans. Two to five plants of each species were infested by Homoptera (mostly scales and membracids) despite the presence of EFNs. Other plant species without EFNs in the vicinity were not found with homopterans more often than the above mentioned species (random samples of up to five nearest neighbouring plants; giving a total of over 200 controls). All homopterans on leaves and stems of the plants were visited by various ant species (Anaplolepis longipes and three Crematogaster spp. were dominant). The ants also visited the extrafloral nectaries on these plants. However, often the ants seemed to prefer honeydew from the homopterans. On a leaf of Croton argyratus, for example, almost 2/3 of all ants visited the scale insects on the leaf lamina and only few took notice of the secreted extrafloral nectar. On Mallotus ants generally favoured membracids despite