CIRCUMNUTATION BEHAVIOR OF AN EXOTIC

American Journal of Botany 87(4): 533 538. 000. CIRCUMNUTATION BEHAVIOR OF AN EXOTIC HONEYSUCKLE VINE AND ITS NATIVE CONGENER: INFLUENCE ON CLONAL MOBILITY KATHERINE C. LARSON Department of Biology, University of Central Arkansas, Conway, Arkansas 7035 USA Virtually all plant parts rotate slightly about a central axis, a movement called circumnutation, but vines show exaggerated circumnutation. This study contrasts circumnutation in two congeneric twining vines, specifically focusing on differences in erect and prostrate shoots, and examines the impact of circumnutation on exploitation of available climbing supports and exploration for more distant ones. Elongating shoots of Lonicera japonica and L. sempervirens growing in a common garden were classed as () erect but not climbing, () climbing on a trellis, or (3) prostrate, and their circumnutation quantified by tracking the compass direction of shoot tips. To quantify the impact of different circumnutation behaviors, the climbing success of erect shoots and the maximum dispersion and rooting success of prostrate shoots were measured. Erect shoots of both species circumnutated at similar rates (averaging 3 /h), and did not differ in their success rate of exploiting trellises (averaging 76.8%). Prostrate shoots differed, with those produced by L. japonica having reduced circumnutation. In contrast, prostrate shoots of L. sempervirens continued to circumnutate as much as erect shoots. The specialized circumnutation behavior of the prostrate shoots of L. japonica results in increased rooting success and maximum dispersion compared to the unspecialized shoots of L. sempervirens. Key words: circumnutation; clonal mobility; introduced species; Lonicera japonica; Lonicera sempervirens; vine. With careful observation, virtually all growing plant parts can be shown to rotate slightly around a central axis, a movement termed circumnutation by Darwin (875). In most plants circumnutation serves no apparent function, occurring as a by-product of growth, and thus accounting for its ubiquitous occurrence in all elongating plant organs (Brown, 993). However, in some plants, such as vines, and in some plant organs, such as stolons, an exaggerated form of circumnutation has evolved that increases the likelihood of encountering a support stem or avoiding obstacles (Darwin, 875, 88). For example, in twining vines, the growing shoot tip is not vertically oriented, but instead bends to the side, while slightly older internodes induce revolutions of the entire young shoot, such that the tip traces an ellipse. When a support host is encountered, continued circumnutation results in the shoot twining up the support. Darwin (875) documented the rates of circumnutation of twining stems, tendrils, and petioles in over 00 species of vine, and in 88 described circumnutation patterns for most parts within nonclimbing plants. While the physiology and biophysics of circumnutation have been recently studied (Johnsson, 979; Silk, 989; Putz and Holbrook, 99; Brown, 993), the significance of variation in the circumnutation behavior of plants in the field has not received much attention. The goal of this paper is to examine variation in circumnutation within and between two related species of twining vine and examine how different circumnutation behaviors influence the vines exploration and exploitation of support hosts. Finding a support host is critical to a climbing vine Manuscript received 7 July 998; revision accepted 9 August 999. The author thanks the University of Central Arkansas University Research Council for funding this project. Carl Dick and Terri Lundberg assisted in the field, Dan Magoulick provided statistical advice, and two AJB reviewers provided helpful comments on the manuscript. 533 (Putz, 984), and vines have mechanisms allowing them to exploit supports within their immediate surroundings as well as explore more distant locations. Depending on the species, circumnutating shoots sweep through areas with diameters ranging from 40 to 00 cm surrounding the main axis and twine up available support hosts (Darwin, 875). Little is known about how specific circumnutation patterns impact exploitation of support hosts. In one study, however, Peñalosa (98) found that morphological specialization in Marsdenia laxiflora, specifically leafless, relatively thin, circumnutating shoots, appeared to increase its attachment success, especially on thinner supports, compared to another species with thicker shoots. Although circumnutation was not specifically measured in this study, the difference in morphology of the two species is likely to influence their circumnutation behavior. Mechanisms to explore more distant locations for support hosts are also important, because most circumnutating shoots fail to encounter a support (three of 64 marked vine shoots successfully climbed a support host; Putz, 984), or do not successfully ascend supports that are found (Peñalosa, 98). In some tropical vines, exploration for more distant supports is accomplished by producing leafless stolons that grow across the forest floor before turning upward and climbing (Hallé, Oldeman, and Tomlinson, 978; Peñalosa, 983, 984). In this study I compare circumnutation in two honeysuckle vines similar in form and distribution in the southeastern USA, but with differing ecologies. Lonicera japonica is a weedy exotic species in the southeastern USA capable of invading and dominating a wide variety of habitats; L. sempervirens is native to the area and is restricted to fewer habitat types where it does not dominate the vegetation (Oosting, 94; Robertson, Robertson, and Tague, 994). Both species produce erect shoots that

534 AMERICAN JOURNAL OF BOTANY [Vol. 87 twine up support hosts as well as prostrate shoots that root where contact with the ground is made and allow lateral spread. While morphological differentiation of erect and prostrate shoot has occurred in many species (Hallé, Oldeman, and Tomlinson, 978; Sachs, 988), the erect and prostrate shoots of these honeysuckle species are morphologically similar despite the different microhabitats they encounter. A previous study of these two congeners indicated that L. japonica allocated biomass differently in climbing shoots and prostrate or trailing shoots, while L. sempervirens allocation patterns were more similar in the two shoot types (Schweitzer and Larson, 999). I hypothesized that in addition to allocation pattern, circumnutation behavior differed between the two species and that differences in circumnutation could contribute to the documented difference in invasiveness of the two species. The questions I address focus on the impact of circumnutation on the ability to climb support hosts and on the ability to locate new support hosts through lateral spread. Specifically, for each species, I asked: () how do circumnutation rates of erect and prostrate shoots differ, () what percentage of erect shoots successfully climb, (3) what is the maximum distance moved by prostrate shoots, and (4) what is the rooting success of prostrate shoots. METHODS I measured shoot circumnutation rate and subsequent shoot climbing success, lateral spread, and rooting success on plants growing in a common garden established in an old field in central Arkansas. The garden was started from rooted cuttings of L. japonica and L. sempervirens taken from plants growing in the Ouachita National Forest, Arkansas. The garden was designed to address additional questions not considered in this paper and contained 8 plants of each species, 4 with a support trellis and 4 without a support trellis. Within this garden, I monitored all shoots that had grown eight nodes by 3 April 997. Any shoots that ceased growing due to damage to the apical meristem were eliminated, leaving a total of 90 shoots of L. japonica and 8 shoots of L. sempervirens. The number of shoots produced by each plant ranged from one to seven, and in this study all shoots of a species are pooled and analyzed as a single population; I do not attempt to examine withinplant and between-plant variation here. The orientation of shoots varied from vertical to horizontal, and each tagged shoot was classified as () climbing if it was twining up the support trellis, () erect if it was not climbing and its angle from horizontal was 5, or (3) prostrate if its angle from horizontal was 5. The climbing success of erect shoots was quantified only for the plants that were growing with an adjacent trellis. Climbing success was measured as the percentage of the total erect shoots produced that were found twining up the trellis by the end of the season. Prostrate shoots quickly grew away from the trellis, and thus had no opportunity to climb. To quantify the ability of the monitored shoots for lateral dispersal, I measured their maximum displacement. Because shoots sometimes turned back on themselves, growing back toward the source plant, I quantified maximum displacement by measuring the distance between the original source plant and the most distant nodes of each monitored shoot, rather than to the actual shoot tip. Circumnutation rates were measured on all tagged shoots by taking readings of the compass direction in which the shoot tip pointed at 3.5- h intervals during a day. The first reading for the day established a reference point, and two additional readings at 3.5-h intervals allowed calculation of compass degrees moved per hour. Preliminary studies examining circumnutation indicated that a complete rotation takes on average 7 h, thus a 3.5-h period reduced ambiguity about whether a TABLE. Two-way ANOVA on the degrees of circular movement (data log transformed) by shoots of L. japonica and L. sempervirens. Shoot types were classed as erect, climbing or prostrate as described in Methods. Source of variation df MS F P Day Species Shoot type Species shoot type Error Day Species Shoot type Species shoot type Error Day 3 Species Shoot type Species shoot type Error 65 65 65.6 83.48 46. 3.9 8.3 0.93.85 3.8 46.4 58.36 7.45 5.64 6.57 5.37 4.0.3 6.93 5.98 8.8 0.34.3 0.0 0.003 0.005 0.70 shoot found pointing in the same compass direction as the previous reading had remained motionless or moved 360. Shoots of Lonicera always rotate clockwise (personal observation; Darwin, 875). Circumnutation for each shoot was measured on three separate days (9,, 7 April). Because a shoot s classification could change from day to day (i.e., an erect shoot could become a climbing shoot or a prostrate shoot), each day s data were analyzed separately rather than as replicated measures. The data were log transformed to meet the assumption of homogeneity of variances and analyzed with a two-way ANOVA, with species (L. japonica or L. sempervirens) and shoot type (climbing, erect, prostrate) as factors. The circumnutation pattern of these shoots was observed by videotaping between two and five erect, climbing, and prostrate shoots of each species. Rooting success of prostrate shoots was quantified for each species. For each node along the entire length of each shoot, I measured the height above the ground and the presence of roots. Because some erect shoots eventually fall and thus secondarily become prostrate, only shoots that had been classed as prostrate on all sample dates were used in this analysis. These data were used to calculate the mean height of nodes, the mean height of rooted nodes, and the percentage of rooted nodes for each shoot. Contrasts between species were made with t tests. RESULTS Lonicera japonica and L. sempervirens produced similar total numbers of primary shoots, with similar percentages of climbing, erect, and prostrate shoots. For example, at the second sampling period on April, of the 54 L. japonica shoots growing with a trellis, 33% were erect but not climbing, 4% were climbing on the trellis, and 6% were prostrate. For the 35 L. sempervirens shoots growing with a trellis, 40% were erect but not climbing, 37% were climbing on the trellis, and 3% were prostrate. Circumnutation rates depended on shoot type (erect, prostrate, or climbing) and on species (Table ). A significant interaction between species and shoot type on day and day indicated that the difference between circumnutation rates of different shoot types was not typically the same for both species and warranted further analysis. I used one-way ANOVA to contrast differences among shoot types for each species, and differences between species for each shoot type (Fig. ).

April 000] LARSON CIRCUMNUTATION IN HONEYSUCKLE VINES 535 Fig.. Circumnutation of shoots of L. japonica and L. sempervirens by shoot type on three days: (a) day, (b) day, and (c) day 3. Significant differences (P 0.05, Tukey) among the different shoot types within each species are indicated by letters above the columns. Significant differences (*P 0.05, **P 0.0, ***P 0.00, t test) between species for each shoot type are indicated above the bar connecting the contrasted pair. Circumnutation rates and patterns were similar on all three days they were measured. Erect shoots showed the greatest circumnutation rates, moving an average of 3 / h. Although the rate of circumnutation by L. sempervirens was slightly greater on all three days, it was never statistically greater than L. japonica (Fig. a c). Climbing shoots had reduced rates of circumnutation in both species, but the reduction was only significant for L. japonica on day (Fig. a c). Overall, the climbing shoots circumnutated at an average rate of 7 /h, with reduced rates expected when the lower internodes are immobilized by contact with the trellis. The significant interaction term in the two-way AN- OVA was due to the different behavior of the prostrate shoots of L. japonica and L. sempervirens. Prostrate shoots on L. japonica had very reduced rates of circumnutation, averaging only. /h, revolving significantly less than erect and climbing shoots of L. japonica, and less than the prostrate shoots of L. sempervirens (Fig. a c). In contrast, the prostrate shoots of L. sempervirens moved at the rate of 5.8 /h, similar to the rates found in climbing shoots. Observations of video-taped shoots indicate that erect shoots of both species rotate in a nearly circular pattern until contact with the trellis is made. Observations on prostrate shoots indicate a striking difference in the pattern of L. sempervirens and L. japonica. Lonicera sempervirens continues to show a circular pattern of circumnutation, however, because the shoots only rotate clockwise, when horizontal, the shoot tip often contacts the ground, resulting in temporary stoppage of the typical rotation. With continued growth and circumnutation, pressure builds up until the tip suddenly springs free. The shoot then resumes circumnutation, but often growing in a different compass direction. Lonicera japonica prostrate shoots circumnutate in a very elongated ellipse, appearing to move up and down only, with little lateral movement. Erect shoots of the two species did not differ in their ability to encounter and climb the support trellises ( 0.48, df, P 0.05). Of the total 8 shoots of L. sempervirens that were monitored, 35 grew adjacent to a trellis and were monitored for climbing success. Twentysix of the 35 shoots adjacent to a trellis were classified as erect at the start of the season, and by the end of the season 9 were climbing on the trellis for a climbing success rate of 73.%. Of the total 90 shoots of L. japonica that were monitored, 54 grew adjacent to a trellis and were monitored for climbing success. Thirty-six of the 54 shoots adjacent to a trellis were erect at the start of the season, and by the end of the season 9 were climbing on the trellis for a climbing success rate of 80.5%. This is consistent with the previous data showing no difference in the circumnutation rates or patterns of the erect shoots of each species. In addition to differing in circumnutation rates, the prostrate shoots of the two species differed in their maximum displacement and rooting success. Maximum displacement was significantly greater for L. japonica than for L. sempervirens, despite a nearly equal total length of prostrate shoots (Fig. ). Maximum displacement of nodes was reduced in L. sempervirens because the shoots would often make abrupt turns, while in L. japonica, shoot growth was almost always in a single compass direction. The rooting success of prostrate shoots of L. japonica was significantly greater than in L. sempervirens. Although L. japonica roots more readily when propagated from cuttings (Schweitzer and Larson, 999), the height of rooted nodes did not significantly differ between species in these field-grown shoots (Fig. 3). All nodes cm above the ground rooted, regardless of species. How-

536 AMERICAN JOURNAL OF BOTANY [Vol. 87 Fig.. Maximum dispersion and total shoot length of 9 L. japonica (open) and 3 L. sempervirens (filled) prostrate shoots. Bars indicate SE, and the P value of a two-sample t test contrasting species is printed above the constrast for each characteristic measured. ever, the percentage of nodes that rooted was over twice as great for L. japonica than L. sempervirens because as prostrate shoots elongated, the nodes of L. japonica remained significantly closer to the ground than in L. sempervirens (Fig. 3). The overall result was that after one growing season, 7 of 9 L. japonica shoots were rooted at one or more nodes, while only 3 of 3 L. sempervirens shoots were rooted at one or more nodes. DISCUSSION Observations of plants made at appropriate time scales reveal circumnutation of shoots on an hourly basis and constant growth into new areas at the apical meristems on a seasonal basis. Plants control their growth into new areas through growth forms that enhance mobility (Callaghan et al., 990; Cain, 990, 994), plastic shoot architecture (de Kroon and Hutchings, 995), and/or division of labor among shoots (Hutchings and Wijesinghe, 997). The results of my study indicate that circumnutation pattern and rates as measured in hours can also significantly impact long-term movement patterns of the plant. While the circumnutation rate and pattern of erect shoots were similar in both species, I found a striking Fig. 3. Height above the ground and percentage of nodes rooted for 9 L. japonica (open) and 3 L. sempervirens (filled) prostrate shoots. Bars indicate SE, and the P value of a two-sample t test contrasting species is printed above the constrast for each characteristic measured. difference in the circumnutation behavior of prostrate shoots of the invasive L. japonica and the native L. sempervirens. Darwin (875) showed that the rate of circumnutation decreases when internodes below the tip are physically restrained from movement, thus it is expected that circumnutation rate would be reduced when more basal internodes are held stationary by either a trellis (i.e., climbing shoots) or the ground surface (i.e., prostrate shoots). The pattern and rate of circumnutation in the prostrate shoots of L. sempervirens appear little modified, except by interference with the ground, from that observed in erect shoots. In contrast, circumnutation was highly modified in the prostrate shoots of L. japonica. Instead of the shoot tip rotating to point in all compass directions during a day, the tip was always found pointing in approximately the same direction, and thus accounting for the very low circumnutation rate recorded. Prostrate shoots of L. japonica circumnutate in the form of a very narrow ellipse with the long axis vertically oriented. The result is a shoot that grows along the ground surface, able to rise over obstacles, but returning to the ground surface again. Reduced circumnutation of prostrate shoots increased the lateral spread of L. japonica in two different ways: increased rooting success and increased maximum displacement. Whether a node rooted or not was dependent on its remaining close to the ground surface ( cm). The increased ability of L. japonica prostrate shoots to root (89.4% of prostrate shoots rooted in their first season of growth) over L. sempervirens (56.5% of prostrate shoots rooted in their first season of growth) is mainly due to the modified circumnutation of L. japonica shoots which keeps nodes close to the ground surface. Within the rooted shoots, less than half as many nodes of L. sempervirens produced roots because circumnutation by these prostrate shoots resulted in a horizontal stem that only made ground contact at a few points, thus reducing the likelihood of rooting. Modified circumnutation also increased the maximum displacement of rooted nodes of L. japonica prostrate shoots. Studies on simulations of clonal plant movement have shown that branching angle can have a strong impact on displacement, i.e., the distance from the source plant to tips of daughter ramets (Cain, 990). Maximum

April 000] LARSON CIRCUMNUTATION IN HONEYSUCKLE VINES 537 speed of movement is attained when daughter ramets are produced on lateral branches with small angles of divergence from the parent shoot (Angevine and Handel, 986; Cain, 994). The reduced circumnutation of L. japonica results in straight growing shoots that are similar to a clonal plant with a branching angle of 0. In the prostrate shoots of L. sempervirens, the ground surface blocks the continued circumnutation of shoot tips around a horizontally oriented axis. Continued growth and circumnutation eventually provide enough pressure to release the shoot, but often with a sharp change in the direction of growth. This appears analogous to larger branching angles in clonal plants and often brings the shoot tip back to the vicinity of the source plant even after growing 3 m in length. By the end of one growing season, the maximum displacement of nodes of prostrate shoots of L. japonica was 39% greater than found in L. sempervirens. Although the prostrate shoots of L. japonica show no morphological specialization, their circumnutation behavior appears similar to the morphologically specialized stolons of three species of clonal herb (Fragaria, Saxifraga, and Cotyledon) described by Darwin (88). He found that the mostly vertical movement of stolons contributed to their ability to move over or around obstacles (Darwin, 88). In contrast, the circumnutation behavior of the prostrate shoots of L. sempervirens does not result in effective stolon-like movement, although they are capable of rooting when nodes are close enough to the ground. The specialized circumnutation behavior of prostrate shoots of L. japonica appears to contribute to its invasiveness by increasing clonal mobility. Although many different factors contribute to the invasiveness of particular species, increased mobility gained through clonal growth form can be important for some species (Stoll, Egli, and Schmid, 998). In addition, the capacity to grow vertically as a vine or clonally across the ground surface contributes to L. japonica s ability to grow in abundance in habitats ranging from old fields to mature forests (Robertson, Robertson, and Tague, 994), while L. sempervirens is not recorded from as many habitat types (Oosting, 94). Clearly no one factor can explain why a particular species may act as a strongly competitive invasive, while its native congener does not, but for these honeysuckle species a greater morphological plasticity (Sasek and Strain, 99; Schierenbeck, Mack, and Sharitz, 994; Schweitzer and Larson, 999) and the production of shoots with specialized circumnutation behavior appear to be major contributors to the invasive nature of L. japonica. It is not clear whether the stolon-like circumnutation pattern found in prostrate shoots of L. japonica is derived from a reduction in circumnutation of erect shoots or whether prostrate stolon-like shoots were already present in a nonclimbing ancestor. The twining habit is rare within the Caprifoliaceae, occurring only in the genus Lonicera and only in a minority of the species within this genus. Most species of Lonicera and most other genera in the Caprifoliaceae are shrubs with an architectural form conforming to the model of Champagnat, defined as producing only orthotropic (vertical) shoots that bend under their own mass to become plagiotropic (horizontal) (Hallé, Oldeman, and Tomlinson, 978). Lonicera sempervirens appears little changed from the basic architectural form of the family, producing only a single type of shoot that varies in orientation from vertical to horizontal. In contrast, the circumnutation behavior of L. japonica has differentiated in two distinct types of shoots. It appears that despite the lack of morphological distinction, L. japonica has evolved to produce distinct orthotropic and plagiotropic branches, a characteristic that appears rarely within the family Caprifoliaceae. Plants gain mobility through their surroundings by circumnutation and morphologically plastic growth, yet despite Darwin s publication of The Power of Movement in Plants in 88, recent studies on plant foraging have only investigated the role of morphological plasticity (Angevine and Handel, 986; Callaghan et al., 990; Cain, 990, 994). My study indicates that a greater understanding of plant foraging emerges when circumnutation is investigated as well. Considering the variation in plant movement documented by Darwin (88), it is likely circumnutation influences plant foraging in other species as well. For example, Evans and Cain (995) studied morphological plasticity and its impact on foraging in a patchy environment, but also found an unexplained ability of the rhizomes of the clonal plant Hydrocotyle to veer away from grass-filled patches. Darwin (88) showed that root tips circumnutate (although the amplitude is small because of the surrounding soil) and respond to environmental cues such as pressure. Darwin demonstrated that a circumnutating root tip would veer away from a higher pressure to the side with lower resistance. Although Evans and Cain (995) were documenting the veering of rhizomes, its likely that similar properties could be responsible. LITERATURE CITED ANGEVINE, M. W., AND S. N. HANDEL. 986. Invasion of forest floor space, clonal architecture, and population growth in the perennial herb Clintonia borealis. Journal of Ecology 74: 547 560. BROWN, A. H. 993. Circumnutations: from Darwin to space flights. Plant Physiology 0: 345 348. CAIN, M. L. 990. Models of clonal growth in Solidago altissima. Journal of Ecology 78: 7 46.. 994. Consequences of foraging in clonal plant species. Ecology 75: 933 944. CALLAGHAN, T. V., B. M. SVENSSON, H. BOWMAN, D. K. LINDLEY, AND B. A. CARLSSON. 990. Models of clonal plant growth based on population dynamics and architecture. Oikos 57: 57 69. DARWIN, C. 875. The movements and habits of climbing plants, nd ed. John Murray, London, UK.. 88. The power of movement in plants. D. Appleton, New York, New York, USA. DE KROON, H., AND M. J. HUTCHINGS. 995. Morphological plasticity in clonal plants: the foraging concept reconsidered. Journal of Ecology 83: 43 5. EVANS, J. P., AND M. L. CAIN. 995. A spatially explicit test of foraging behavior in a clonal plant. Ecology 76: 47 55. HALLÉ, F., R. A. OLDEMAN, AND P. B. TOMLINSON. 978. Tropical trees and forests: an architectural analysis. Spinger-Verlag, Berlin, Germany. HUBER, H., A. FIJAN, AND H. J. DURING. 998. A comparative study of spacer plasticity in erect and stoloniferous herbs. Oikos 8: 576 586. HUTCHINGS, M. J., AND D. K. WIJESINGHE. 997. Patchy habitats, division of labour and growth dividends in clonal plants. Trends in Ecology and Evolution : 390 394. JOHNSSON, A. 979. Circumnutation. In W. Haupt and M. E. Feinleib

538 AMERICAN JOURNAL OF BOTANY [Vol. 87 [eds.], Encyclopedia of plant physiology, vol. 7, 67 646. Springer- Verlag, Berlin, Germany. OOSTING, H. J. 94. An ecological analysis of the plant communities of the Piedmont, North Carolina. American Midland Naturalist 8: 6. PEÑALOSA, J. 98. Morphological specialization and attachment success in two twining lianas. American Journal of Botany 69: 043 045.. 983. Shoot dynamics and adaptive morphology of Ipomoea phillomega (Vell.) House (Convolvulaceae), a tropical rainforest liana. Annals of Botany 5: 737 754.. 984. Basal branching and vegetative spread in two tropical rain forest lianas. Biotropica 6: 9. PUTZ, F. E. 984. The natural history of lianas on Barro Colorado Island, Panama. Ecology 65: 73 74., AND N. M. HOLBROOK. 99. Biomechanical studies of vines. In F. E. Putz and H. A. Mooney [eds.], The biology of vines, 73 97. Cambridge University Press, Cambridge, UK. ROBERTSON, D. J., M. C. ROBERTSON, AND T. TAGUE. 994. Colonization dynamics of four exotic plants in a northern Piedmont natural area. Bulletin of the Torrey Botanical Club : 07 8. SACHS, T. 988. Ontogeny and phylogeny: phytohormones as indicators of labile change. In L. D. Gottlieb and S. K. Jain [eds.], Plant evolutionary biology, 57 8. Chapman and Hall, London, UK. SASEK, T. W., AND B. R. STRAIN. 99. Effects of CO enrichment on the growth and morphology of a native and introduced honeysuckle vine. American Journal of Botany 78: 69 75. SCHIERENBECK, K. A., R. N. MACK, AND R. R. SHARITZ. 994. Effects of herbivory on growth and biomass allocation in native and introduced species of Lonicera. Ecology 75: 66 67. SCHWEITZER, J. A., AND K. C. LARSON. 999. Greater morphological plasticity of exotic honeysuckle species may make them better invaders than native species. Journal of the Torrey Botanical Society 6: 5 3. SILK, W. K. 989. On the curving and twining of stems. Environmental and Experimental Botany 9: 95 09. STOLL, P., P. EGLI, AND B. SCHMID. 998. Plant foraging and rhizome growth patterns of Solidago altissima in response to mowing and fertilizer application. Journal of Ecology 86: 34 354.