Northeastern Naturalist 520 J.R. Milks, J. Hibbard, and T.P. Rooney 22001177 NORTHEASTERN NATURALIST 2V4(o4l). :2542,0 N–5o2. 54 Exfoliating Bark Does Not Protect Platanus occidentalis From Root-climbing Lianas James R. Milks1, Justin Hibbard1, and Thomas P. Rooney1,* Abstract - Lianas are structural parasites that depress growth, fertility, and survival rates of their hosts, but the magnitude to which they alter these rates differ among host species. We tested the hypothesis that Platanus occidentalis (Sycamore) would have fewer adventitiousroot climbing lianas than other tree species. We reasoned that because Sycamore possesses exfoliating bark, it would periodically shed newly-established lianas from the trunk. We investigated the distribution of lianas on the trunks of trees ≥10 cm DBH in floodplains in southwestern Ohio. Contrary to our predictions, Sycamore trees had significantly more lianas than expected at 3 of 5 sites, and significantly fewer than expected at 1 site. In contrast, Acer negundo (Boxelder) had less than half the lianas expected. We find no support for our hypothesis that bark exfoliation protects Sycamore trees from climbing lianas, and suggest possible mechanisms that might protect Box Elder from adventiti ous-root climbing lianas. Introduction Lianas are woody vines that use tree hosts for structural support. They often depress the growth, fertility, and survival rates of their hosts (Givnish 1992, 1995; Ingwell et al. 2010; Ladwig and Meiners 2009; Stevens 1987; van der Heijden and Phillips 2009). Tree species have evolved different mechanisms to decrease the number of lianas that successfully become established on their trunks. These mechanisms include trunk spines, guarding ants, flexible trunks, compound leaves, long leaves, and high relative-growth rates (Givnish 1995; Putz 1980, 1984). Putz (1984) reported higher liana mortality rates on trees with trunk spines. Guarding ants have been observed removing lianas from host trees, e.g., Azteca ants living in the hollow internodes of Cecropia trees (Janzen 1973, Putz 1980). When lianas grow from one tree canopy to another, flexible trunks can facilitate dislodging the connection when trees sway in opposite directions (Putz 1984). Lianas attached to long leaves or compound leaves become dislodged as those leaves are shed (Putz 1980, 1984). Trees with high relative-growth rates are more effective in breaking twining lianas than slower-growing trees (Putz 1980). The few studies that have examined the role of bark shedding as a defense against lianas (e.g. Carsten et al. 2002. Sfair et al. 2016, Talley et al. 1996a) have been confined to tropical species. Bark shedding would be expected to protect against liana infestation because lianas would be shed along with pieces of bark. This mechanism should be especially effective against lianas that climb via adventitious roots (hereafter, root-climbing lianas) because they attach to bark to climb. 1Department of Biological Sciences, Wright State University, Dayton, OH 45435. *Corresponding author - thomas.rooney@wright.edu. Manuscript Editor: Elizabeth N. Hane Northeastern Naturalist Vol. 24, No. 4 J.R. Milks, J. Hibbard, and T.P. Rooney 2017 521 Talley et al. (1996a) noted that bark shedding reduced lianas in 2 species of Australian rainforest trees. Carsten et al. (2002) found a more complex pattern—liana densities increased at intermediate levels of bark shedding but decreased at higher levels of shedding. Sfair et al. (2016) found that trees with exfoliating bark did not have fewer lianas, but species with exfoliating bark combined with other anti-liana mechanisms had fewer lianas. Temperate floodplains in the eastern US are well suited for studying liana– host relationships. Floodplain forests are subject to several factors that increase liana abundance including disturbance through periodic flooding (van der Heijden and Philips 2008) and forest fragmentation (Londré and Schnitzer 2006). Floodplains are also the primary habitat of Platanus occidentalis L. (Sycamore), a bark-shedding deciduous tree in the eastern US (Burns and Hon kala 1990). Although bark-shedding has been hypothesized to protect Sycamores from lianas (Givnish 1992, 1995), to our knowledge, no previous studies have tested this hypothesis for Sycamore. In this study, we tested the hypothesis that a temperate-zone bark-shedding tree, Sycamore, would have fewer root-climbing lianas than co-occurring species that do not shed bark. We counted the number of root-climbing lianas on tree trunks in 5 floodplain forests in southwestern Ohio, and we predicted that Sycamore would have fewer lianas than expected compared to non-bark–shedding species. Field-site Description We conducted this study in mature floodplain forests at 5 different parks in southwestern Ohio (39.7ºN, 84.1ºW): Germantown, Huffman Dam, Sugarcreek, and Taylorsville Metroparks in Montgomery County, and The Narrows Preserve in Greene County. Montgomery County parks occur within the Great Miami River Watershed and The Narrows Preserve lies within the Little Miami River Watershed; both drain into the Ohio River. All sites consist of mature, secondary forest. Their origin dates to the 1910s when, following the Great Dayton Flood of 1913, land was acquired for flood-control purposes (Morgan 1951). Forest cover extends less than 200 m perpendicular to the river corridor at all sites. Land use in both watersheds is predominantly cultivated cropland. Forest cover, pasture, and urban development are also present. Both watersheds are located within the Till Plains region of Ohio. This glaciated landscape contains rolling hills, moraines, and outwash plains (Zimmerman and Runkle 2010). Floodplain forests are comprised of mature deciduous species. Sycamore and Boxelder were the dominant tree species at our study sites. The invasive shrub Lonicera maackii (Rupr.) Herder (Amur Honeysuckle) is common in the forestshrub layer (Hutchinson and Vankat 1998). Methods We recorded the diameter at breast height (DBH) and species of each tree ≥10 cm DBH in a single 10 m x 300 m belt transect (total 0.30 ha) in mature floodplain Northeastern Naturalist 522 J.R. Milks, J. Hibbard, and T.P. Rooney 2017 Vol. 24, No. 4 forests at each park, and considered each park as a separate site. We placed transects randomly within the forests, but in most cases, they were within 50 m of forest edges due to the narrow dimensions and fragmented nature of the floodplain forests in this region. We tallied the number of root-climbing lianas present on the trunk of each tree at 1.6 m above ground level. We did not attempt to distinguish between separate individuals. Instead, we counted the number of stems present in a band around the tree at the1.6-m sampling height. We chose root-climbing lianas because we predicted that they would be susceptible to being shed by trees with exfoliating bark. We collected data in the spring for 2 field seasons (2007 and 200 8). We determined mean (± SE) number of lianas per tree, importance values (IV), and expected numbers of lianas per tree species for each site. We calculated importance values for each tree species by adding the relative DBH and relative density of each species, then dividing by 2 and multiplying by 100. We calculated relative DBH by dividing total DBH for each species by total DBH for all trees per site and relative density by totaling all individual stems per species and dividing by the total number of individual stems per site. If lianas were randomly distributed among trees, we would expect the number of lianas per tree species to be distributed in proportion to the IV of each tree species. Importance values incorporate both the age/size and density; thus, tree species with higher IVs would host more lianas than tree species with smaller IVs due to increased surface area to which lianas could attach (Buron et al. 1998, Carsten et al. 2002, Leicht-Young, et al. 2010, Reddy and Parthasarathy 2006, Talley et al. 1996a). We calculated the expected number of lianas per tree species as the product of the number of lianas at a site and the tree species IV at the site, divided by 100. We employed replicated goodness-of-fit (G) tests with Williams correction for continuity to analyze differences between observed and expected lianas per tree species of the 2 species with the highest IV, Sycamore and Acer negundo L. (Box Elder), with each site considered a replicate (Sokal and Rohlf 1981). Calculations were performed in Microsoft Excel 2008 for Mac and validated using manual calculations. Results We measured 1145 trees comprising 18 species and counted 1417 root-climbing lianas (mostly Toxicodendron radicans (L.) Kuntze [Poison Ivy] and a few Parthenocissus quinquefolia (L.) Planch.) [Virginia Creeper]) in a total of 1.5 ha. Of the 18 tree species encountered, Sycamore and Box Elder had the highest IVs: 31.2 and 25.6, respectively. The remaining 16 species had a combined IV of 43.2. We found 568 lianas (40.0%) growing on Sycamore. This number was significantly greater than expected at 3 of 5 sites, significantly less at 1 site, and did not differ from expected abundance at the remaining site (Table 1). When we pooled the data across sites, Sycamore had 33% more root-climbing lianas than expected (pooled G = 20.5, df = 1, P < 0.001). In contrast, we found only 142 (10.0%) lianas on Boxelder, which had significantly fewer lianas than expected at 3 sites and Northeastern Naturalist Vol. 24, No. 4 J.R. Milks, J. Hibbard, and T.P. Rooney 2017 523 did not differ from expected abundance at 2 sites (Table 1). When we pooled data across sites, Box Elder had 55% fewer root-climbing lianas than expected (pooled G = 67.7, df = 1, P < 0.001). Discussion We found no support from our data for the hypothesis that bark-shedding protects Sycamore from root-climbing lianas. Sycamores had either the same as or more than the expected number of lianas at 4 sites out of 5 sites, whereas we predicted that the species would have fewer than expected lianas. This result contrasts with Talley et al. (1996a), who found that bark-shedding trees in Queensland, Australia, tropical forests had fewer than expected root-climbing lianas. Carsten et al. (2002) found that root-climbing lianas increased on trees with intermediate bark roughness and levels of bark-shedding and decreased at high levels of shedding and on trees with smooth bark. It is possible that Sycamore falls within the intermediate range of the bark-texture scale of Carsten et al. (2002). One possible test would be to compare individual Sycamore trees for differences in bark-shedding levels and liana loads because individual Sycamores vary in levels of bark shedding, with some trees shedding nearly all bark and others shedding very little (J.R. Milks, pers. observ.). Alternatively, exfoliating bark alone may be an ineffective anti-liana mechanism (Sfair et al. 2016). In contrast to Sycamore, Boxelder had either the expected number of lianas or significantly fewer lianas than expected. Other studies have also noted fewer than expected lianas on the closely related Acer saccharum Marsh. (Sugar Maple). Both Talley et al. (1996b) and Leicht-Young et al. (2010) found fewer than expected Poison Ivy lianas on Sugar Maple in forests in Alabama, Indiana, and Michigan. Possible reasons for the differences between liana abundance between Sycamore and Boxelder include leaf size, bark morphology, and bark chemistry. Putz Table 1. Number of tree stems, importance values (IV) per species per site, observed liana abundance, expected liana abundance, and G-values for goodness of fit. For G-values, ns = not significant, *P less than 0.05, **P < 0.005. Observed # Expected # Species/site # of stems IV per site of lianas of lianas G-value Acer negundo (Box Elder) Germantown 120 58.4 11 13.4 1.0 ns Huffman 94 26.8 44 169.7 163.8** Narrows 88 20.6 75 108.4 14.5** Sugarcreek 6 4.2 0 7.1 14.4** Taylorsville 38 14.3 12 18.3 2.7 ns Platanus occidentalis (Sycamore) Germantown 24 18.3 5 4.2 0.2 ns Huffman 36 18.0 143 114.0 8.5** Narrows 105 37.8 269 175.0 77.9** Sugarcreek 50 46.7 127 79.4 55.1** Taylorsville 77 41.0 24 52.5 28.9** Northeastern Naturalist 524 J.R. Milks, J. Hibbard, and T.P. Rooney 2017 Vol. 24, No. 4 (1984) found that trees with leaves >50 cm in length protected trees from lianas on Barro Colorado Island, Panama. Plants that shed such large leaves are more likely to dislodge attached lianas, compared to plants with smaller leaves (Putz 1984). Although Sycamore generally has larger leaves than Boxelder, most are less than 25 cm in length, making leaf size an unlikely mechanism of liana shedding in eastern temperate-floodplain forests. Bark morphology (smooth versus furrowed) is also unlikely to be an important mechanism, because this factor has been tested in other forest types with mixed results (Boom and Mori 1982, Carsten et al. 2002). In our study, Boxelder had slightly furrowed bark. Bark morphology by itself is unlikely to explain our results, although it warrants further study. One unexplored possibility is that allelopathic chemicals in the bark of some maple species may protect them from root-climbing lianas. Talley et al. (1996b) found that allelopathic chemicals in Sugar Maple bark (as well as chemicals in the bark of other tree species) could inhibit liana seedling germination and growth in the southern US, and differences in the presence of allelopathic chemicals influenced liana distribution on host trees. Talley et al. (1996a) found similar patterns in Australia. It is possible that bark chemistry may also protect maple species from clinging lianas, although our study did not investigate that possibility. This study is, to our knowledge, the first to demonstrate that bark shedding in Sycamore does not protect that species from liana infestation. We also showed that Boxelder has either the same or fewer than expected number of lianas, which is also a new finding. Future investigations could examine host preferences for different species of lianas in temperate floodplains, and whether variability in bark shedding among individual Sycamore individuals affects liana loads. Acknowledgments The idea for this project was first suggested by T. Givnish. We thank Mike Bottomley for statistical consulting, and 2 reviewers for constructive comments in an early draft of this manuscript. Raw data from this study can be accessed at https://figshare.com/articles/ VineShedding_SW_Ohio_csv/3125773. Literature Cited Boom, B.M., and S.A. Mori. 1982. Falsification of 2 hypotheses on liana exclusion from tropical trees possessing buttresses and smooth bark. Bulletin of the Torrey Botanical Club 109:447–450. Burns, R.M., and B.H. Honkala. 1990. Silvics of North America: 2. Hardwoods. Agriculture Handbook 654. US Department of Agriculture Forest Service, Washington, DC. 877 pp. Buron, J., D. Lavigne, K. Grote, R. Takis, and O. Sholes. 1998. Association of vines and trees in second-growth forest. Northeastern Naturalist 5:359–36 2. Carsten, L.D., F. A. Juola, T.D. Male, and S. Cherry. 2002. Host associations of lianas in a southeast Queensland rainforest. Journal of Tropical Ecology 18:107–120. Givnish, T.J. 1992. Nature green in leaf and tendril. Science 256:1339–13 41. Givnish, T.J. 1995. Plant stems: Biomechanical adaptation for energy capture and influence on species distributions. Pp.. 3–49, In B.L. Gartner (Ed.). Plant Stems: Physiology and Functional Morphology. Academic Press, San Diego, CA. 460 pp. Northeastern Naturalist Vol. 24, No. 4 J.R. Milks, J. Hibbard, and T.P. Rooney 2017 525 Hutchinson, T.F., and J.L. Vankat. 1998. Landscape structure and spread of the exotic shrub Lonicera maackii (Amur Honeysuckle) in southwestern Ohio forests. American Midland Naturalist 139:383–390. Ingwell, L.L., S.J. Wright, K.K. Becklund, S.P. Hubbell, and S.A. Schnitzer. 2010. The impact of lianas on 10 years of tree growth and mortality on Barro Colorado Island, Panama. Journal of Ecology 98:879–887. Janzen, D.H. 1973. Dissolution of mutualism between Cecropia and its Azteca ants. Biotropica 5:15–28. Ladwig, L.M., and S.J. Meiners. 2009. Impacts of temperate lianas on tree growth in young deciduous forests. Forest Ecology and Management 259:195–200. Leicht-Young, S.A., N.B. Pavlovic, K.J. Frohnapple, and R. Grundel. 2010. Liana habitat and host preferences in northern temperate forests. Forest Ecology and Management 260:1467–1477. Londré, R.A., and S.A. Schnitzer. 2006. The distribution of lianas and their change in abundance in temperate forests over the past 45 years. Ecology 87:2 973–2978. Morgan, A.E. 1951. The Miami Conservancy District. McGraw-Hill, New York, NY. 504 pp. Putz, F.E. 1980. Lianas vs. trees. Biotropica 12:224–225. Putz, F.E. 1984. How trees avoid and shed lianas. Biotropica 16:19–23. Reddy, M.S., and N. Parthasarathy. 2006. Liana diversity and distribution on host trees in 4 inland tropical dry evergreen forests of peninsular India. Tropical Ecology 47:109–123. Sfair, J.C., A.L.C. Rochelle, A.A. Rezende, J. van Melis, R.J. Burnham, V. de L. Weiser, and F.R. Martins. 2016. Liana avoidance strategies in trees: Combined attributes increase efficiency. Tropical Ecology 57:559–566. Sokal, R.R., and F.J. Rohlf. 1981. Biometry, 2nd Edition. W.H. Freeman and Company, New York, NY. 859 pp. Stevens, G.C. 1987. Lianas as structural parasites: The Bursera simaruba example. Ecology 68:77–81. Talley, S.M., W.N. Setzer, and B.R. Jackes. 1996a. Host associations of 2 adventitiousroot– climbing vines in a north Queensland tropical rain forest. Biotropica 28:361–366. Talley, S.M., R.O. Lawton, and W. N. Setzer. 1996b. Host preferences of Rhus radicans (Anacardiaceae) in a southern deciduous hardwood forest. Ecolog y 77:1271–1276. Van der Heijden, G.M.F,. and O.L. Phillips. 2008. What controls liana success in Neotropical forests? Global Ecology and Biogeography 17:373–383. Van der Heijden, G.M.F. and O.L. Phillips. 2009. Liana infestation impacts tree growth in a lowland tropical moist forest. Biogeosciences 6:2217–2226. Zimmerman, C.L., and J.R. Runkle. 2010. Using ecological land units for conservation planning in a southwestern Ohio watershed. Natural Areas Journal 30:27–38.