sign in sign up
<<
Home , Petiole (botany), Polybia occidentalis

University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

Faculty Publications in the Biological Sciences Papers in the Biological Sciences

11-1977

The Extrafloral Nectaries of Ipomoea carnea (Convolvulaceae)

Kathleen H. Keeler University of Nebraska-Lincoln, [email protected]

Follow this and additional works at: https://digitalcommons.unl.edu/bioscifacpub

Part of the Commons

Keeler, Kathleen H., "The Extrafloral Nectaries of Ipomoea carnea (Convolvulaceae)" (1977). Faculty Publications in the Biological Sciences. 275. https://digitalcommons.unl.edu/bioscifacpub/275

This Article is brought to you for free and open access by the Papers in the Biological Sciences at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Faculty Publications in the Biological Sciences by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. Keeler in American Journal of Botany (November-Decemeber 1977) 64(10). Copyright 1977, Botanical Society of America. Used by permission.

Amer. J. Bot. 64(10): 1182-1188. 1977.

THE EXTRAFLORAL NECTARIES OF IPOMOEA CARNEA (CONVOLVULACEAE)l

KATHLEEN H. KEELER School of Life Sciences, University of Nebraska, Lincoln 685~8

A B S T R A C T lpomoea carnea (Convolvulaceae) possesses two types of extrafloral nectaries, located on the petiole and on the . These secrete a complex containing and amino acids. The attracted to the extrafloral nectaries are predominantly and they are relatively abundant throughout the year. A number of incidents of defense as a result of the presence of extrafloral nectary visitors at the extrafloral nectaries of I. carnea were observed and are consistent with the -guard theory of the function of extrafloral nectaries.

RECENTLY, increased attention has been given to Elias, 1972; Carroll, 1974; Keeler, 1975b). Re- the question of the function of extrafloral nec- cent work has for the most part supported the taries and particularly to the hypothesis that the idea that the function of extrafloral nectaries is ecological function of these nectaries is to attract to attract insects, especially ants, which defend ants which protect the plant (Elias, 1972; Elias the plant from herbivores. Elias' morphological and Gelband, 1975; Elias, Rozich, and New- work on Pithecellobium macradenium (Legu- combe, 1975; Bentley, 1976; 1977). In this minosae) (1972) and Turnera ulmifolia (Tur- paper the activity of the extrafloral nectaries of neraceae) (Elias et al., 1975) indicated that the a tropical dry forest , Ipomoea carnea (Con- extrafloral nectaries of these two are very volvulaceae), is described and evidence is given complex structures with nectar production cycles that supports the contention that extrafloral nec- correlated with the herbivores are expected taries function to attract insects which are pro- to be the most active, which would support their tective to the plant. role in an ant-plant association. Extrafloral nectaries are glands which secrete Elias and Gelband (1975) also found that the an aqueous solution of and other chemicals four extrafloral nectary systems of Campsis rad- (Baker and Baker, 1973) that can be found on icans () are responsible for attract- aboveground parts of plants outside the . ing ants and that in the presence of abundant ants, They occur in a variety of monocotyledonous and there is a decrease in the frequency of flower rob- dicotyledonous families (Zimmermann, 1932; bing.2 Bentley (1976) found that a significant Schnell, Cusset, and Quenum, 1963). The mor- amount of the variance in the percentage of phology and chemical composition of extrafloral that matured on Bixa oreliana (Bixaceae) was nectaries have been analysed in a number of cases. explained by a linear regression of the abundance Nevertheless, the function of extrafloral nectar- of ants at the site. She (1977) also showed that ies is not well established. when artificial extrafloral nectaries were created In the case of ant-inhabited , extrafloral on common bean seedlings (Phaseolus vulgaris), nectaries have been shown to provide the major by placing karo syrup on the , the treated source of sugar for the resident ant population plants weighed significantly more at the end of (Janzen, 1966, 1967; Brown, 1960). However, the experiment than did the control plants in the where the insects which visit extrafloral nectaries rainy season. On the other hand, Lukefahr and do not reside on the plant, the story is less clear others (1960; 1965; 1966) have shown that and has been debated for over a century (see strains of nectariless cotton, derived from crosses of Gossypium tomentosum, had significantly 1 Received for publication 10 February 1977; revision lower infestations of a number of important pest accepted 8 April 1977. species, under a range of field and cage condi- I wish to thank H. G. Baker, I. Baker, P. A. Opler than did and Y. B. Linhart for their advice and encouragement. I tions, extrafloral nectary-bearing cotton am grateful to the help of the Costa Rican staff of the strains. Organization for Tropical Studies (O.T.S.), especially J. In this paper observations and experiments on Campabadal. T. S. Elias, R. B. Kaul and R. W. Shuel the extrafloral nectaries of Ipomoea carnea Jacq. made helpful comments on the manuscript. This research in Costa Rica was funded in part by NSF GB-237256, NSF GB- (Convolvulaceae) are reported 255924, NIH GM-367 and an O.T.S. student subsistence grant. This work was done in partial fulfillment of the 2Van der Pijl (1954) observed the presence of ants to Ph.D. degree, Department of Genetics, University of disturb and displace flower-robbing Xylocopa spp. bees California, Berkeley. in Java. 1182 November-December, 1977] KEELER-EXTRAFLORAL NECTARlES OF IPOMOEA 1183

I S .: r # _ 'Gi, : ,~~~~~~~~~~~~~~~~~ _-g .....r 3if

Fig.~~*_1, 2. .:Cmoou bretes takin exrfoa neta fro eioanetr of Ipomoea...... cre. X 1.....;|S

2~~.Acuulto of neca at peiela exrfoa netre of I. carnea. .X.

which support the work of Bentley, Elias, and season, because, as the dry season proceeds,Ipo- their predecessorsin suggesting that the extra- moea carnea loses its leaves. Flowering is con- floralnectaries attract protective ants to the plant. fined to the dry season, so the pedicellar ex- trafloral nectaries function only at that . MATERIALS AND METHODS-Ipomoea carnea Consequently,there is little overlap between the Jacq. [syn. I. fistulosa, (Austin, in press)] is a period of petiolar and pedicellarextrafloral nec- twining woody vine which is sufficiently robust tary activity. to supportitself to a height of two meters. It is The function of floral nectaries is well estab- found on both the Atlantic and Pacific coasts of lished to be part of the process of ,so Mexico south to Argentina (Matuda, 1964; the activityof the floral nectariesof I. carnea will O'Donell, 1952). The populationstudied was in not be considered here. However, it should be GuanacasteProvince, Costa Rica, near the Or- pointed out that althoughthe plant is a melitto- ganizationfor Tropical Studies'Palo Verde field philous obligate outcrosser (Keeler, 1975a, b), station. This area is in the tropical dry forest the flower-visitinginsects are entirely different zone of Holdridge (1967), receiving 1,800 mm from the insects which visit the extrafloralnec- of rainfall a year, but having a distinct dry sea- taries. son betweenDecember and May. The chemical constitutentsof the were determinedby Irene Baker using spot tests and Ipomoea carnea was chosen for study because chromatography(Keeler, 1975b). The frequency of the abundant nectar it produced. It was and distributionof the extrafloralnectary visitors thought that the function of nectaries would be were studied in a series of surveysin which were more easily demonstratedin a plant which pro- recordedthe insects visiting the nectaries of 500 duced copious quantitiesof nectar. In addition, or 1,000 branches. Two different branches per 1. carnea was abundantnear the Palo Verde field plant were observed whenever possible. Two station and consequently more accessible than such surveys were made in the rainy season many other extrafloralnectar producing plants. (October and November, 1973) and one in the I. carnea has two extrafloralnectary systems. dry season (January,1974), in the pasturestudy An extrafloralnectary is located on either side of areaat Palo Verde. the distal end of the petiole. A ring of five specimens were identified as follows: extrafloralnectaries occurs at the base of each pseudomyrmecineants by D. H. Janzen, Dept. flower around the pedicel. The petiolar extra- of Zoology, University of Michigan; other ants floral nectaries are most active during the rainy by R. R. Snelling,Los Angeles County Museum; 1184 AMERICAN JOURNAL OF BOTANY [Vol. 64

TABLE 1. Free sugars in Ipomoea carnea nectarsa TABLE 2. Amino acids in Ipomoea carnea nectarsa

Nectar Extrafloral Floral Nectar Petiolar Pedicellar Extrafloral Floral Amino acid Petiolar Pedicellar Sucrose 33 8b 35.8 54.8 Glucose 32.9 33.1 29.1 Alanine 0.0062a 0 0.076 Fructose 24.2 29.1 16.1 Arginine 0.0291 0.0342 0.131 Melibiose 3.9 trace 0 Asparagine 0.0012 0.0616 0.057 Raffinose 5.9 0 0 Glutamic acid 0.0029 0.0027 0.001 a Determined by chromatography using solvent of Glycine 0 0 0.006 Smith (1969), 316 and stain of Saini (1966). Isoleucine 0.0069 0 0 Serine 0.0518 0.0383 0 b Percent of total sugar. a Determined by chromatography of dansylated amino acids and fluorometry. Method of I. Baker (unpubl.). larval lepidoptera by T. Davies, California Acad- Determined by I. Baker. emy of Sciences, San Francisco; and other in- b ,m/ml. sects by P. A. Opler, Office of Endangered Spe- cies, U.S. Fish and Wildlife Service, Washington, D.C. much longer than that of petiolar extrafloral nec- taries, which increased the difference in total nec- RESULTs-Extrafloral nectaries-The petiolar tar production compared to the rainy season. extrafloral nectaries are functionally mature be- The sugar and amino acid contents of Ipomoea fore the lamina mature (Fig. 1). First nectar is carnea nectaries are given in Tables 1 and 2. The observed when the is about 4 cm long and floral nectar content is included for comparison. nearly ready to unfold. The nectaries continue In addition, by using spot tests of Baker and to produce nectar as the leaf expands to its ma- Baker (1973) floral and extrafloral nectars of ture size (roughly cordate, about 15 cm at its Ipomoea carnea were found to contain antioxi- greatest length and width). The nectaries func- dants and small amounts of phenolics, aldehyde tion for a few days after full size is attained and and ketones. Unsaturated lipids were absent then gradually become inactive. Their total ac- from the extrafloral nectars but were present in tivity lasts from two to three weeks, and the rate small quantities in the floral nectar. No alkaloids of nectar production is about 0.2 1l per leaf per or proteins were detected in any of the three nec- day. tars. Pedicellar extrafloral nectaries develop with Variation in the nectar produced by the dif- the flower buds (Fig. 2). They begin secreting ferent types of nectaries was tested. Nectars from nectar when the buds first separate from the clus- the petiolar nectaries on different individual plants ter. Production continues at an average rate of were more like each other than like those of the about 0.2 ,d per flower per day throughout the pedicellar nectaries for total amino acid content. period of development of the , opening of In 16 petiolar extrafloral nectaries from as many the flower, and until the is dry and ready plants, the average amino acid content (ninhy- to dehisce. This process takes four to six weeks. drin scale of Baker and Baker, 1973) was found Nectar is produced by night as well as by day to be 1.9 (range, 0 to 3.0). Pedicellar extra- in both petiolar and pedicellar nectaries. No floral nectar from 26 nectaries averaged 2.6 changes in rate were detected. However, the rate (range, 1.0 to 3.5). (Because of the different of production per nectary is low enough to make times of the year that the extrafloral nectaries accurate measurements difficult and it is being are active, these were not collected simultaneous- further investigated. ly.) The means were significantly different from Although no daily changes in extrafloral nec- each other (P < 0.025). tar production were detected, there was great seasonal variation. Plants produced much more Nectary visitors-The extrafloral nectary visi- nectar in the dry season than in the rainy season. tors would be relevant to the function of extra- Although the rate of nectar production on buds, floral nectaries if they act as ant-guard attractants. and was found to be about the same As was stated, there was no overlap between as on the leaves (about 0.2 1I per day), a branch floral nectary and extrafloral nectary visitors. A which had five nectary-bearing leaves in the rainy wide variety of insects was observed visiting the season produced on the average four to eight clus- extrafloral nectaries of Ipomoea carnea. ters of flowers, each bearing ten to thirty buds. The percent of extrafloral nectaries with a nec- Thus, there was a ten- to fifty-fold increase in the tar-feeding insect present (but not necessarily number of extrafloral nectaries on a plant in the taking nectar) when approached was found to be dry season. In addition, as indicated, the func- 4.5% in October, 3.6% in November and 0.26% tional life of pedicellar extrafloral nectaries was in January. These figures actually represent a November-December, 1977] KEELER-EXTRAFLORAL NECTARIES OF IPOMOEA 1185 much higher rate of visitation of the nectaries by TABLE 3. Flower robbing in Ipomoea carnea at Palo insects since both flying and crawling insects Verde, Guanacaste Province, Costa Rica. Jan., 1974a tend, if undisturbed, to visit every nectary on a branch before leaving. Therefore, in the rainy Flowers season, 89.8% and 73.0% of the branches had at Number Percent least one nectary-visiting insect present and in the robbed robbed Area dry season 15.5% of the flowering branches had Experiment I a visitor, when encountered, suggesting consider- Low-visitor area 40 69 58 Mb able activity at any particular nectary. High-visitor area 36 116 31** NC Of the insects which visited Ipomoea carnea Experiment II extrafloral nectaries, the majority in all surveys Low-ant plants 112 171 66 M were ants. In October, 89.5% of the 2,761 in- High-ant plants 11 36 31** M sects were ants. In November, 63.7% of the 546 insects seen were ants. (This survey covered 503 a See text for details. branches, compared to 1,000 in the other sur- b M, Midpasturearea. veys.) In January, ants made up 59.2% of the N, Area north of pasture. 191 insects seen visiting nectaries. (A fourth sur- **, Significantly different from control at 1% level. vey, in January, 1974 in an adjacent pasture with taller vegetation revealed that ants were 91.2% of the 522 insects observed.) In terms of both of extrafloral nectary visitors resulted in reduc- absolute and relative numbers, ants were the most tion of damage to the plant. At the Palo Verde abundant extrafloral nectary visitors. study site, Xylocopa frontalis robbed 50-70% of Compared to other insects, ants were seen near the Ipomoea carnea flowers each day. These bees the nectaries for sustained periods of time. In are too large to enter the flower legitimately. 518 min of observation of extrafloral nectaries in Instead, they climb over the lamina and walk periods of 20 to 30 min, one or more ants were down the corolla tube on the outside toward the present 55.3% of the time, while flies (miscel- base of the flower. There, they pierce a hole laneous Diptera) were present only 29.9% of the in the corolla and take nectar without pollinat- time. The other visitors, largely Coleoptera and ing the flower. In doing so they stand close to , were collectively present 36.8% of the pedicellar extrafloral nectaries. the time (individuals of some additional species The results of a series of observations of the were present for an entire observation period but relationship of flower robbing to activity in the only occurred in one of the 22 observation area of the extrafloral nectary are given in Table periods; those are not included here). The vari- 3. For plants in an area of relatively high extra- ance in these estimates, which resulted from the floral nectary visitor density (present at 31 % of the different behaviors of the species included in branches) 36 of 116 flowers were robbed (31%), each group, precludes statistical testing. A full compared to 40 of 69 flowers (59% ) in an area description of extrafloral nectary visitors seen at of lower visitor frequency (15% of the branches L. carnea is given in Keeler (1975b). with a visitor). This difference is statistically sig- The most important species of ants visiting nificant (P < 0.01). Furthermore, on plants Ipomoea carnea extrafloral nectaries were: For- within the low-density area that were consistently micinae: Camponotus abdominalis, C. substitu- covered with ants (Camponotus brettesi and tus, C. brettesi, and C. rectangularis; Myrmicinae: Crematogaster ampla), presumably due to prox- Crematogaster ampla, Monomorium ebininum, imity to sites, the frequency of robbing, over Solenopsis geminata, and S. littoralis; Pseudo- several days, was 31%, compared to 66% for myrmicinae: Pseudomyrmex gracilis; and Doli- randomly selected flowers from the same field. choderinae: Iridomyrmex pruinosum. Ant fre- Solenopsis geminata were observed to be an quencies varied greatly between surveys. In all effective ant-guard in an unusual situation. S. three surveys the genera Solenopsis and Cam- geminata were foraging at the extrafloral nec- ponotus were the most numerous, making up taries of a small (1 m) L. carnea plant, which 30.7% and 24.6% of all ants observed. Of the was growing in contact with a small (2 m) six species in these two genera, only Camponotus of collinsii Stafford which was occupied brettesi was at all consistent in its frequency- by Pseudomyrmex ferruginea ants. As described comprising 22.6% of the ants in the first survey, by Janzen (1966, 1967) P. ferruginea will de- 22.9% in the second, and 43.5% in the third. The fend A. collinsii from herbivores and will also other species showed great variation, including kill parts of other plants that come in contact absence from the third survey (in which only 191 with it. In the Palo Verde study area, I. carnea insects were seen in 1,000 observations). usually suffered great damage from P. ferruginea attack when it was next to an occupied acacia. Evidence of ant-guarding-Observations were In this case, however, the presence of the Solenop- made consistent with the theory that the activities sis workers effectively reduced the damage by 1186 AMERICAN JOURNAL OF BOTANY [Vol. 64

Fig. 3, 4. 3. Results of contact of lpomoea carnea with Pseudomyrmex ferruginea-inhabited Acacia collinsii: tip and several leaves have been chewed off by acacia-ants. X 1/2. 4. Effectiveness of foraging by Solenop- sis geminata in reducing damage due to P. ferruginea. Although the I. carnea plant shown here was growing next to the same A. collinsii shown in Fig. 3, the damage was much less. Note that the shoot tip and all leaves are still in- tact. X 1/2.

Pseudomyrmex workers. Pseudomyrmex work- reestablishedin all directions. Subsequentanaly- ers crossed onto 1. carnea and attackedit fiercely sis of the data did not provide any convincing at the petioles. Whenevera Solenopsisworker ap- evidence that some of the plants had different peared, however, Pseudomyrmexthat had been natural levels of ant density, and no differences attacking the Ipomoea petiole would climb up in level of damage were seen between the differ- onto the lamina. If it was actually encountered ent plants. by a Solenopsis,the latter would become excited, An example of predationon an L. carnea her- to the point of the Pseudomyrmex. This bivore by an extrafloralnectary visitor was ob- rarely occurred,since the Pseudomyrmexants al- served. A occidentalis was cap- most invariablyretreated as soon as a Solenopsis tured on a leaf of L. carnea with its prey, a small approached. The result was that the Pseudo- larva of the Lymantriidaeor Liparidae,one of the myrmexants did little damage. They chewed on known herbivoresof I. carnea [specimensUniv. the leaf surface,but muchless photosyntheticsur- of Californiasurvey #190376 (P. occidentalis) face was lost than from the usual cutting through and prey 190680-692 (larva lepidoptera)]. P. of the petiole. Nor were they able to stay on the occidentalis made up 0.3-2.0% of the nectary- apical meristemlong enough to sever it. In two visitinginsects observed. weeks, there was little damage to the I. carnea visited by Solenopsis geminata except for pieces DIsCUSSION-Thesignificance of the particular taken out of the leaf margins, while a similar combinationof chemicalsin the nectar is not ob- sized L. carnea on the other side of the same aca- vious at present. In other studies of the content cia, which was not visited by Solenopsis,lost five of extrafloralnectaries, Butler et al. (1972) found leaves and the apical (Fig. 3, 4). Thus a range of proportionsof fructose, sucrose and ant-guardingcan be effective, even against as se- glucose in the extrafloral nectaries of cotton vere an "herbivore" as Pseudomyrmex ferruginea. (Gossypium barbadense and G. hirsutum). In A series of experiments aimed at excluding all cases they found sucrose to have the lowest ants and testing the resultinglevels of damageto proportion(5-24% of the nectar), with fructose the leaves failed because it proved impossible to and glucose about equal in amount. Bowden exclude ants. The vine grew as much as 5 cm per (1970) found the extrafloralnectar of the Ni- internode per day amid tall vegetation and con- gerian grass Andropogon gayanus to contain 42% tact with the surroundingplants was continuously sucrose, 27% glucose, and 24% fructose, 1% November-December, 1977] KEELER-EXTRAFLORAL NECTARIES OF IPOMOEA 1187

each of arabinose, xylose, and maltose, less than activity on the plant. On the other hand, Luke- 1 % raffinose and 3 % of three unidentified sugars. fahr and others (1960, 1965, 1966) found Thus the glucose, fructose and sucrose contents of greater damage by pest insects to cotton plants I. carnea extrafloral nectaries differ from those with extrafloral nectaries than to those without, found in others species, but are similar to them. under a range of field conditions. The cause of The presence of raffinose in the petiolar extra- the differences in results for cotton has not been floral nectar (Table 1) is interesting in view of the demonstrated, but it is likely that the explana- fact that Frisch (1971) reported raffinose to be tion lies in the environment-that the agricultural highly attractive to ants but not to honeybees. conditions under which cotton is raised are suf- This suggests an added attractant for ants in the ficiently different from 4he natural habitats of the petiolar extrafloral nectar. Why raffinose is ab- other studies to produce a different result. In par- sent from the pedicellar extrafloral nectar requires ticular, cultivation methods (plowing, tilling, ap- further investigation. plication of insecticides) may reduce ant popula- Fewer amino acids are found in I. carnea extra- tions sufficiently to destroy their effectiveness. floral nectar than in cotton (Gossypium hirsutum) In any event, the results of this study support an (Hanny and Elmore, 1974) but all of the ones ant-guard interpretation of the function of extra- present in I. carnea (Table 2) are also present floral nectaries in Ipomoea carnea. in cotton. The function of amino acids in extra- floral nectar is unknown. In floral nectar they LITERATURE CITED are believed to be additional attractants to pol- AUSTIN, D. F. In press. Ipomoea carnea Jacq. vs. 1. linators because of their nutritional value (Baker fistulosa Mart. ex. Choisy. Taxon. and Baker, 1975). BAKER, H. G., AND I. BAKER. 1973. Amino acids in If extrafloral nectaries function to attract ant- nectar and their evolutionary significance. Nature guards, a selective advantage must exist. Ant- 241: 543-545. guards at the nectaries must convey an advantage , AND . 1975. Studies of nectar-constitu- for the plant in terms of increased sexual or asex- tion and -plant coevolution, p. 100-140. ual . In this study the advantage has In L. E. Gilbert and P. H. Raven [ed.], Coevolution not been proven, but it is suggested by the protec- of and plants. Univ. Texas Press, Austin. BENTLEY, B. L. 1976. Plants bearing extrafloral nec- tion afforded I. carnea from flower robbers and taries and the associated ant community: inter- acacia-ants by activity at the plant's extrafloral habitat differences in the reduction of herbivore nectaries. damage. Ecology 57: 815-820. In addition, two necessary but not sufficient 1977. The protective function of ants visiting conditions for ant-guarding to exist on a plant are the extrafloral nectaries of Bixa orellana (Bixaceae). the presence of abundant ants at the nectaries and J. Ecol. 65: 27-38. aggressive behavior by the ants. The presence of BOWDEN, G. D. 1970. The sugars in the extrafloral abundant ants was demonstrated in the nectary- nectar of Andropogon gayanus var. bisquamula- visitor surveys. Many of the ant species observed tus. 9: 2315-2318. BROWN, W. J. 1960. Ants, acacia and browsing mam- are highly aggressive around a food source (Wil- mals. Ecology 41: 587-593. son, 1971). BUTLER, G. D., JR., G. M. LOPER, S. E. McGREGOR, J. Aggressive behavior by the ants feeding at Ipo- L. WEBSTER, AND H. MARGOLIS. 1972. Amounts moea carnea extrafloral nectaries was demon- and kinds of sugars in the nectaries of cotton (Gos- strated in the situation where Solenopsis gemi- sypium spp.) and the time of their secretion. Agron. nata foragers displaced Pseudomyrmex ferruginea J. 64: 364-368. workers attacking I. carnea. on herbi- CARROLL, B. L. 1974. Extrafloral nectaries: adapta- vores of the plant was observed in this case for tions to reduce herbivore damage. Ph.D. disserta- hymenoptera (Polybia occidentalis), but not for tion, University of Kansas, Lawrence. ants. the conditions for ant- ELIAS, T. S. 1972. Morphology and anatomy of foliar Thus, necessary nectaries of Pithecollobium macradenium (Legu- guarding exist at the extrafloral nectaries of Ipo- minosae). Bot. Gaz. 133: 38-42. moea carnea and, indeed, protection by ant- , AND H. GELBAND. 1975. Nectar: its produc- guards has been seen under some conditions. tion and function in trumpet creeper. Science 189: The results of this study are consistent with 289-290. those of Bentley (1976, 1977), who demon- , W. R. ROZICH, AND L. NEWCOMBE. 1975. The strated ant-guarding at the extrafloral nectaries foliar and floral nectaries of Turnera ulmifolia L. of Bixa orellana (Bixaceae) and at artificial "ex- Amer. J. Bot. 62: 570-576. trafloral nectaries" on Phaseolus vulgaris (Legu- FRISCH, K. VON. 1971. Bees, their vision, chemical minosae). Taylor and Inouye (unpubl.) found senses and language. Revised ed. Cornell Univ. a reduction in in Heli- Press, Ithaca, N.Y. significant predation HANNY, B., AND C. D. ELMORE. 1974. Amino acid anthella quinquenervis (Compositae) in plants composition of cotton nectar. J. Agric. Food Chem. with higher ant densities. Elias and Gelband 22: 99-104. (1975) observed decreased frequency of flower HOLDRIDGE, L. R. 1967. Life zone ecology. Tropical robbing in Campsis radicans with increased ant Science Center, San Jose, Costa Rica. 1188 AMERICAN JOURNAL OF BOTANY [Vol. 64

JANZEN, D. H. 1966. Coevolution of mutualism be- MATUDA,E. 1964. El genero Ipomoea en Mexico. An. tween ants and acacias in Central America. Evolu- Inst. Biol. Mex. 35: 49-50. tion 20: 249-275. O'DONELL, C. A. 1952. Nota sobre Ipomoea fistulosa 1967. The interaction of the bull's horn aca- Martius ex Choisy. Bol. Soc. Argent. Bot. 4: 175- cia (A. cornigera L.) with one of its ants inhabitants 176. (Pseudomyrmex ferruginea F. Smith) in eastern PIJL, L. VAN DER. 1954. Xylocopa and flowers in the Mexico. Kansas Univ. Sci. Bull. 47: 315-558. tropics. K. Nederlandse Ak. Waters. Proc. 57, ser. KEELER, K. H. 1975a. Ipomoea carnea in Costa Rica. C: 413-423, 542-562. Brenesia 5: 1-6. SAINI, A. S. 1966. Some technical improvements in . 1975b. Plant adaptation: insect defense and the paper chromatographyof sugars. J. Chromatogr. enzyme variation. Ph.D. dissertation. Univ. of 24: 484-486. Calif., Berkeley. SCHNELL, R., G. CUSSET, AND M. QUENUM. 1963. LUKEFAHR, M. J., C. B. COWAN, T. R. PREMMER, AND Contributiona l'etude des glandes extra-florales chez L. W. NOBLE. 1966. Resistance of experimental quelques groupes de plantes tropicales. Rev. Gen. cotton strain 1514 to the bollworm and cotton flea- Bot. 70: 269-313. hopper. J. Econ. Entomol. 59: 393-395. SMITH, I. 1969. Chromatography and electrophoretic - D. F. MARTIN, AND J. R. MEYER. 1965. Plant techniques. 3rd ed. Interscience Publ., New York. resistance to five lepidoptera attacking cotton. J. WILSON, E. 0. 1971. The insect societies. Belknap Econ. Entomol. 58: 516-518. Press of Harvard Univ., Cambridge, Mass. , AND C. RHYME. 1960. Effects of nectariless ZIMMERMANN,J. 1932. Uber die extraflorale Nec- cottons on populations of three lepidopterous in- tarien der Angiospermen. Bot. Centr. Beih. 49: sects. J. Econ. Entomol. 53: 242-244. 99-196.