PROPAGATION AND TISSUE CULTURE

HORTSCIENCE 48(12):1525–1529. 2013. Axenic plantlets grown on synthetic media provide optimal experimental control over most abiotic and biotic parameters. However, Germinating and Culturing Axenic because T. radicans has no economic value other than being a noxious invasive , Poison Ivy Seedlings there is rather limited knowledge about 1 T. radicans germination patterns. Two prior Elise B. Benhase and John G. Jelesko studies suggest that T. radicans drupes re- Plant Pathology, Physiology, and Weed Science Department, Virginia Tech, quire scarification to initiate seedling germi- 548 Latham Hall, Blacksburg, VA, 24061-0390 nation. One study focused on T. radicans seed dispersal by birds and squirrels demon- Additional index words. dermatitis, drupe, endocarp, mesocarp, tissue culture, strated that sandpaper scarification significantly radicans, urushiol increased seedling germination frequencies Abstract. Urushiols are the chemical constituents responsible for causing the character- relative to untreated drupes (Penner et al., istic skin rash resulting from contact with poison ivy [ subsp. 1999). A different study oriented on T. radi- radicans (L.) Kuntze] plant tissue. Future detailed physiological and molecular studies cans interactions with host tree species used of T. radicans urushiol metabolism will require the production and cultivation of axenic a combination of physical (pounding) and T. radicans in controlled environments. To this end, the present study focused on chemical (sulfuric acid) treatments to obtain treatments to enhance germination and reduce microbial contamination to obtain axenic adequate seedling germination frequencies T. radicans seedlings. Toxicodendron radicans drupes treated singularly with water, to support studies into the effects of host tree bleach, cold, or gibberellic acid showed very low germination frequencies. In contrast, allelochemical effects on T. radicans seedling concentrated sulfuric acid strongly promoted seedling germination by removing exo- germination and growth (Talley et al., 1996). carp, mesocarp, and causing pitting of the brachysclereid and osteosclereid layers of the These limited T. radicans findings are endocarp. Most T. radicans drupes harbored significant amounts of fungal and bacterial consistent with more extensive seed germi- contaminants. Although the serial mechanical scarification, sulfuric acid, and bleach nation studies of closely related Rhus species treatments promoted seedling germination, this serial treatment regime was not (). Untreated drupes from five adequate to render the majority of drupes microbe-free. Nevertheless, ’25% of treated Rhus species (, R. typhina, R. T. radicans drupes were axenic, and these needed to be promptly separated from adjacent virens, R. aromatica,andR. trilobata) show fungal-contaminated drupes to avoid cross-contamination. The isolated axenic T. very low seedling germination rates during radicans drupes germinated at high frequency producing viable seedlings that grew well permissive germination conditions (Li et al., in sterile plant culture conditions. 1999c). However, drupes from all five spe- cies showed significantly increased germina- tion (albeit to differing degrees) after sulfuric acid scarification. The five Rhus species dif- Poison ivy is best known for its ability to and Dawson, 1953, 1954). Urushiol triolefin fered in whether the embryo dormancy was cause irritating skin rashes called Rhus der- congeners correlate with increased severity enforced by physiological or physical mecha- matitis. Poison ivy belongs to the family of contact dermatitis symptoms compared with nisms. Only R. aromatica showed physiolog- Anacardiaceae, which includes other species those with less unsaturation (Johnson et al., ical dormancy that was broken by gibberellic producing sap capable of causing skin re- 1972). The principal urushiol congeners in acid treatment (Li et al., 1999b). In contrast, actions, including poison oak (formerly Rhus T. radicans are the pentadecylcatechols. Al- treatments that disrupted the physical integ- toxicodendron diversilobum), poison though the chemical composition of T. rad- rity of the endocarp were sufficient to induce (formerly Rhus toxicodendron vernix), icans urushiol and clinical immunology of seedling germination in the other four Rhus (Mangifera indica), cashew (Anacardium occi- the delayed contact dermatitis is well docu- species, indicating physically enforced seed dentale),andtheAsianlactree(Rhus vernici- mented, urushiol physiology and metabolism dormancy (Li et al., 1999c). In the case of flua). Gillis proposed a systematic revision of in T. radicans plants are poorly understood. R. glabra, physical seed dormancy is main- poison ivy, poison oak, and poison sumac Urushiol levels and composition vary in tained by the water-impermeable endocarp from the genera Rhus to Toxicodendron (Tox- poison ivy plants. In one report, young , tissue, in particular the outermost brachyscler- icodendron radicans, Toxicodendron diversilo- fresh young stems, fruits, and bark showed eid and internally proximal osteosclereid cell bum,andToxicodendron vernix, respectively) high urushiol levels, and the triolefin congener layers (Li et al., 1999a, 1999c). The water- (Gillis, 1971). Despite the moniker ‘‘poison’’ comprised over half of the total urushiols impermeable brachysclereid and osteoscler- ivy, the manifested dermatitis is an immuno- present (Craig et al., 1978). A different study eid layers are sensitive to disruption by acid logically based allergic reaction, delayed con- examining leaves of different ages also showed treatment, thereby allowing water to penetrate tact hypersensitivity, not an acute toxicity or high urushiol levels in young leaves with lower the underlying macrosclereid layer resulting in poisoning (Kurtz and Dawson, 1971). urushiol levels in the oldest leaves (Baer et al., embryo imbibition and dormancy release. The natural product responsible for induc- 1980). In the latter study, the diolefin was the Given our interest to investigate urushiol ing the delayed contact dermatitis is generically most abundant congener, whereas the triole- metabolism and physiology during sterile called urushiol. Urushiol refers to a number of fin was the least abundant. Both reports used tissue culture conditions in the future, it is pentadecylcatechol or heptadecylcatechol con- tissues obtained from unmanaged T. radicans currently necessary to develop foundational geners with varying degrees of unsaturation plants in which genetic, abiotic, and biotic methods to germinate and culture axenic ranging from one to three double bonds (Symes factors were neither determined nor controlled. T. radicans seedlings. To this end, the primary From these two studies it is clear that urushiol objective of the present study was to identify composition and levels are not determinate physical, chemical, and cultural treatments traits in poison ivy. Therefore, similar to suitable for producing axenic T. radicans seed- Received for publication 10 Sept. 2013. Accepted other defensive plant secondary metabolites, lings cultured on synthetic media. for publication 16 Oct. 2013. T. radicans urushiol levels and composition J.G.J. is grateful to Roger Harris and Alex Niemiera are likely to change in response to develop- for sharing their insights and enthusiasm for this Materials and Methods nascent poison ivy research project. We also appre- mental, environmental, and biotic factors. As ciate the helpful comments from R.H. and Richard a prerequisite for future detailed urushiol Toxicodendron radicans drupe collection. Veilleux during the manuscript internal review. physiological and metabolism studies, it will Two lianas of T. radicans subsp. radicans 1To whom reprint requests should be addressed; be necessary to grow T. radicans plants under were the source of drupes used in this study. e-mail [email protected]. well-controlled environmental conditions. Drupes from the RoaCo-1 liana, located in

HORTSCIENCE VOL. 48(12) DECEMBER 2013 1525 Catawba, VA, at GPS coordinates latitude subjected to a 30-min 50% bleach treatment then subsequently treated with sterile water, 3722#57.19$ N, longitude 806#43.07$ W, before plating with the goal of drupe surface cold (2 to 6 weeks at 4 C), 1 mg·mL–1 were harvested in August of 2012. Drupes sterilization. The initial drupe treatment exper- gibberellic acid (GA), or SA treatments. from the Huckleberry-1 liana, located on the iments used 50 treated drupes on 100 · 15-mm One-way ANOVA results indicated that both Huckleberry Trail in Blacksburg, VA, at GPS petri plates containing 0.5 · MS basal salts solid replication and sandpaper scarification were coordinates latitude 3712#7.29$ N, longidue media. After initial chemical treatments, the not significant factors (F = 0.45, P = 0.50; and 8024#900$ W, were harvested on 1 Sept. seeds were stored at room temperature in the F = 0.75, P = 0.53, respectively), so these two 2012. The panicles were separated from the dark for 7 d and then transferred to a Percival parameters were removed from subsequent lianas, leaves removed, placed in black CU-36l4 growth chamber (Percival Scientific, ANOVA analyses. Fig. 1 illustrates T. radi- 113.5-L plastic bags in an outdoor shed for Perry, IA) set for constant 28 C and 16 h light. cans seedling germination frequencies in 3 d, and then moved to an air-conditioned Seedling germination was scored 4 weeks post- response to these treatments. Relative to research facility to allow the panicles and drupe treatment. Germination was defined water-treated controls, in which 1% and 3% drupes to fully dry. minimally as the emergence of the radical. In germination frequencies were observed for Drupe scarification treatments. Toxico- subsequent experiments comprised of sulfuric Huckleberry-1 and RoaCo-1 ecotypes, re- dendron radicans drupes were subjected to acid and bleach treatments, only 20 drupes were spectively, there were no significant increases several types of scarification. Initially drupes placed on larger 150 mm · 10-mm petri plates in seedling germination frequencies as a result were removed from the panicles and manu- with solidified 0.5 · MS media. Drupe sterility of either cold or GA treatments. In contrast, ally scarified between two blocks of 60 grit was initially assessed on Day 4 posttreatment SA treatment resulted in significantly greater sandpaper for 1 min. In later experiments, using a Leica Zoom 2000 illuminated stereo T. radicans seedling mean germination fre- drupes attached to panicles were mechani- microscope (20· magnification) examining quency (11% for Huckleberry-1 and 42% for cally scarified by placing them into a ‘‘3 lb closed petri plates. Drupes lacking visible RoaCo-1) relative to water-treated controls rock tumbler’’ (Chicago Electric Power Tools, fungal hyphae and/or bacterial growth were (ANOVA, Tukey correction T = –3.349, P < Chicago, IL) together with nine small rocks scored as sterile drupes. Individual sterile 0.019; and T = –9.756, P < 0.0001, respec- and 50 mL of all-purpose fine sand (Quickrete, drupes were transferred to 0.5 · MS basal salt tively). When factored overall treatments, the Atlanta, GA) and continuously tumbled for a media in either Magenta boxes or Phytatray II RoaCo-1 drupes consistently showed greater duration spanning four nights. The manual and boxes (Sigma-Aldrich Co.) in a laminar flow seedling germination frequencies than mechanically scarified seeds were separated sterile cabinet. At 10 d posttreatment drupes Huckleberry-1 (ANOVA, Tukey correction, from detached exocarp, mesocarp, stem tissue, were scored for seedling germination. Drupes T = 3.81, P = 0.0003), indicating ecotype- sand, and rocks using a No. 25, 710-mmmesh were imaged using a Zeiss Stemi, SV11 Apo specific differences. screen (W.S. Tyler, Mentor, OH) and stored at (Zeiss, Thornwood, NY) dissecting micro- Extensive microbial contamination. All room temperature in paper bags. Chemical scope fitted with a Syncroscopy (Synoptics drupe treatments resulted in substantial mi- scarification comprised treating no more than Inc., Frederick, MD) digital camera. Axenic crobial contamination. The treated drupes 100 seeds in 20 mL of 13 N sulfuric acid (SA) individual seedlings either in Magenta boxes typically showed extensive fungal hyphal (Fisher Scientific, Waltham, MA) in a 50-mL or Phytatrays continued to grow well for growth and, to a much lesser degree, bacterial BD Falcon tube (Becton Dickenson, Franklin several weeks. The sulfuric acid–bleach serial growth. The microbial contamination was Lakes, NJ) placed in a horizontal position treatment experiments were replicated three particularly acute in the 4 C-treated plates, on an orbital shaker at 100 rpm for 30 min at times, each on a different day. All other which showed increasing amounts of fungal room temperature. During the SA treatment, treatments were replicated four times with each growth on plates with longer durations of the tubes were vortexed for 15 s at 10-min replication initiated on a different day. The cold treatment. Extending the duration of 50% intervals during the 30-min acid treatment. resulting data were analyzed using a General bleach treatment did not result in any consis- The SA was removed and the drupes washed Linear Model analysis of variance (ANOVA) tent improvement in fungal surface steriliza- three times with 25 mL sterile distilled water with Tukey correction, or t test using Minitab tion (data not shown). Nevertheless, bleach per wash in a laminar flow sterile cabinet Version 14 (State College, PA) using a # 0.05. treatment was advantageous. Drupes treated (Labconco, Kansas City, MO). only with sterile water showed extensive Initial chemical and temperature drupe Results bacterial contamination with little evidence treatments. Drupes were treated with 50% of fungal contamination, suggesting that bac- commercial liquid bleach, 3% sodium hypo- Sulfuric acid promoted seedling germination. terial growth suppressed fungal growth. Com- chlorite final, (Clorox Co., Oakland, CA), and T. radicans drupes were subjected to four pared with water treatment alone, bleach washed with sterile water as described pre- treatments often used to promote seedling treatment resulted in considerably less bac- viously for the SA treatment. Bleach pre- germination. Drupes from two T. radicans terial contamination such that fungal con- treated seeds were incubated with 10 mL lianas (Hucklebrry-1 and RoaCo-1) were tamination then predominated the observed 1mg·mL–1 gibberellic acid G3 (Alfa Aesar, either unscarified or sandpaper-scarified and microbial contaminants. Interestingly, the Ward Hill, MA) for 0.5 h and then washed with three rinses of 25 mL sterile distilled water. For the cold treatments, bleach-treated drupes were incubated at 4 C for 2, 4, or 6 weeks before transfer to a growth chamber for seedling germination. The cold-treated seeds were assessed for seedling germination after 4 weeks in a long-day growth chamber. Seedling germination conditions. All treated T. radicans drupes were placed on sterile 0.5 · Murashige and Skoog (MS) basal salts media (Murashige and Skoog, 1962) solidi- fied with 0.3% w/v Phytagel (Sigma-Aldrich Co., St. Louis, MO) in plastic petri plates (either 50 · 100 mm or 150 · 100-mm petri plates depending on the experiment). Because the overarching objective of this study was to Fig. 1. Poison ivy seedling germination frequency in response to various treatments. Shaded columns obtain sterile T. radicans seedlings, unless indicate drupes from lianas of Huckleberry-1 (light gray) or RoaCo-1 (dark gray). Gibberellic acid explicitly stated, treated drupes were also (G3) concentration was 1 mg·mL–1. Error bars indicate SE; n = 4 independent replications per mean.

1526 HORTSCIENCE VOL. 48(12) DECEMBER 2013 harsh SA treatment was not effective at eliminating fungal contamination (data not shown). In summary, neither bleach nor SA treatment alone was effective in producing axenic drupes. Optimal mechanical and chemical scarification treatments. Several aspects of the initial seedling germination studies might have contributed to the high levels of fungal contamination. The sandpaper scarification generally removed the brittle exocarp but left much of the mesocarp tissue intact. This was particularly the case for small drupes inter- spersed with larger drupes, the latter of which prevented the sandpaper from coming into Fig. 2. Mean germination frequency of RoaCo-1 poison ivy drupes after pairwise sequential treatments direct contact with the smaller drupes. The involving combinations of water, concentrated sulphuric acid (SA), and bleach (3% sodium failure to efficiently remove mesocarp tissue hypochlorite). Pairwise treatment order is listed from left to right. Error bars indicate SE;n=3 meant there was little abrasion of the endo- independent replications per mean. carp tissues. In general, drupes with more intact mesocarp typically showed greater amounts of fungal contamination, suggesting bleach (SA, bleach) germinated at a mean buoyant drupes (one-way ANOVA, F = 0.26, endophytic fungi residing within the meso- frequency of 0.67 compared with those P = 0.78). However, Fig. 3B illustrates there carp. Another disadvantage of sandpaper treated with bleach and then SA, which showed was a dramatic difference in seedling mean scarification included the occupational expo- a germination frequency of 0.47 (ANOVA, germination frequencies between non-buoyant sure of researchers to fine particulates gen- Tukey correction, T = 3.51, P = 0.03). In these and buoyant drupes (ANOVA, Tukey correc- erated during the ‘‘sanding’’ of poison ivy sequential dual treatments, most (though not tion T = 8.928, P < 0.0001). None of the drupes. Lastly, the drupe plating density on all) of the drupes displayed evidence of fungal buoyant drupes germinated nor did they show standard sized petri plates combined with and, to a much lesser degree, bacterial con- evidence of embryo imbibition. Thus, buoy- the 4-week incubation period before scoring tamination. These results together with the ant drupes most likely contained inviable seedling germination might have obscured aforementioned results demonstrated that embryos even before treatments were applied. the presence of initially sterile drupes that concentrated SA alone, bleach alone, nor se- The non-buoyant drupes contained viable were subsequently infected by adjacent quential SA and bleach treatments were embryos and produced germinating seed- fungal-contaminated drupes on the solid me- adequate to efficiently surface-sterilize most lings. Fig. 3B shows that initially scored dia. Therefore, several refinements in both drupes. However, it was also noted that after sterile (i.e., microbe-free) drupes resulted drupe scarification and cultural methods were a few days, some drupes were apparently not in a mean germination frequency of 0.80 ± made to minimize the presence of mesocarp contaminated with fungi or bacteria. These 0.07 SE. The contaminated drupes showed and minimize fungal cross-contamination later observations led to additional cultural a reduced mean germination frequency of between plated drupes. To obtain more effi- practices used to identify and isolate axenic 0.59 ± 0.05 SE. Although there were consis- cient and even abrasion of exocarp and T. radicans seedlings. tent differences in the mean germination mesocarp tissues, we mechanically scarified Axenic seedling isolation by cultural frequency between sterile and contaminated T. radicans RoaCo-1 drupes attached to methods. To promote the identification and drupes, the observed difference in means panicles by using a combination of small isolation of sterile T. radicans drupes, we was not statistically significant (paired t test, rocks and fine sand in a rock tumbler for either modified or introduced several cultural T =–2.68,P = 0.075). The majority of initially 96 h. This mechanical scarification resulted practices. Mechanically scarified RoaCo-1 T. scored sterile drupes resulted in axenic poi- in drupes that were separated from the pan- radicans drupes were serially treated with SA son ivy seedlings that grew well in Magenta icles, more evenly scarified, and much less and then bleach to promote efficient seedling boxes or Phytatrays for several weeks, as occupational exposure to airborne fine par- germination. All treated drupes were plated evidenced by the production of many lateral ticulates. Interestingly, the sand acquired at a lower density (20 drupes on larger 150 · roots, several true leaves, and the absence of a novel gray color after mechanical drupe 10-mm 0.5· MS plate). Approximately one- visible microbial contamination. Although scarification that differed from both the init- fourth of the treated drupes were buoyant the majority of initially scored sterile drupes ial sand color and the color of T. radicans during the sterile water washes, and these produced fully axenic seedlings, some re- drupes. Likewise, the mesocarp of mechan- were plated separately from the non-buoyant sulted in fungal-contaminated seedlings that ically scarified drupes displayed a gray color drupes. After four nights of incubation in the typically manifested as hyphae-growing within and had a more friable texture that was dif- dark, each drupe was visually inspected using the 0.5· MS media associated with the roots. ferent from both non-scarified and sandpaper- a dissecting microscope and scored as either The frequency of initially scored sterile drupes scarified drupes. sterile or contaminated. At this time, many that produced fungal-contaminated seed- The mechanically scarified T. radicans drupes showed evidence of embryo imbibi- lings was relatively low at 0.15 ± 0.05 SE. drupes were subjected to combinatorial tion, which was comprised of increased drupe We could not determine whether these fungal- bleach and SA treatments to identify optimal size and often cracking of the endocarp. contaminated seedlings were the result of initial conditions for initiating seedling germina- Apparent sterile drupes were removed from inaccurate assessment of sterility or the result tion. Drupes first treated with sterile water the large 0.5· MS petri plates and transferred of secondary fungal contamination occurring and then subsequently with bleach [Fig. 2 to 0.5 · MS media in Magenta boxes or during the transfer of a sterile drupe from the (water, bleach)] showed a comparable low Phytatrays and placed in a long-day growth petri plate to a Magenta box or Phytatray. In seedling mean germination frequency like chamber. The plates containing the contam- summary, the serial process of mechanical in Fig. 1 (water), demonstrating that bleach inated drupes were also transferred to the scarification, SA then bleach chemical treat- treatment did not promote seedling germina- same growth chamber. After 10 additional ments followed by early visual scoring and tion. Again, SA significantly increased seed- days of incubation, both the Magenta boxes separation of microbe-free drupes produced ling germination (Fig. 2). The order of bleach and petri plates were scored for seedling many axenic T. radicans seedlings that grew and SA treatment had a small but significant germination. Fig. 3A illustrates that there well in tissue culture conditions. effect on seedling germination frequency. was no significant difference in mean drupe T. radicans embryo dormancy is maintained Drupes sequentially treated with SA and then sterility frequency between non-buoyant and by enforcing embryo desiccation. Embryo

HORTSCIENCE VOL. 48(12) DECEMBER 2013 1527 parameters as possible. Growing axenic T. radicans plants in sterile plant culture media would provide well-controlled condi- tions in which to characterize urushiol metab- olism. Toward this end, the primary objective of this study was to establish a reliable pro- tocol for efficient poison ivy seedling germi- nation and cultivation in axenic conditions. In the present report, sandpaper scarifica- tion did not increase T. radicans seedling germination frequency. This is in contrast to a previous report that showed significantly increased T. radicans germination rates re- sulting from sandpaper scarification (Penner et al., 1999). Efficient T. radicans seedling germination was achieved by combining me- chanical scarification (rock tumbler, rocks, and sand) with sequential SA and bleach treatments. Mechanical scarification removed all of the brittle exocarp and much mesocarp tissue, including most of the apparent resin ducts. The prior removal of these outer drupe tissues made the SA-mediated removal of the remaining friable mesocarp more efficient, because SA-treated drupes without mechani- cal scarification would often still retain meso- carp tissue on the drupe. With nearly all/most mesocarp removed, the exposed endocarp was therefore more susceptible to SA-mediated dissolution of brachysclereid and osteoscler- Fig. 3. Microbial contamination and seedling germination patterns of poison ivy drupes after rock/sand eid cell layers (i.e., endocarp ‘‘pitting’’). The scarification, sulphuric acid (SA), and bleach treatments. (A) Frequency of sterile/microbe-free drupes. stimulation of T. radicans embryo imbibition Total indicates combined buoyant drupes and non-buoyant drupes. (B) Germination frequency and seedling germination by SA treatment is of drupes of differing buoyancy and differing state of microbial contamination: contaminated similar to some Anacardiaceae members such drupes (light gray) and sterile drupes (dark gray). Error bars indicate SE; n = 4 independent replications per mean. as Rhus glabra in which embryo dormancy is enforced by water-impermeable brachy- sclereid and osteosclereid cell layers that dormancy can be enforced by either physio- seedling typically emerged. In conclusion, maintain embryo desiccation and hence logical or physical mechanisms. Gibberellin based on both the ineffectiveness of GA to dormancy (Li et al., 1999a, 1999b, 1999c). treatment will frequently break physiologi- promote seedling germination together with Thus, T. radicans embryo dormancy is largely cally enforced embryo dormancy. However, the effectiveness of SA treatment to erode enforced by physical isolation of the embryo 1mg·mL–1 GA treatment did not significantly the brachysclereid and osteosclereid cell from environmental moisture. increase T. radicans mean seedling germina- layers, T. radicans drupes maintained embryo The stimulation of T. radians seedling ger- tion frequency (Fig. 1) relative to water treat- dormancy mostly by physical, not physiolog- mination by strong acid treatment may have ment (ANOVA, Dunnett test, T = –0.52, P = ical, mechanisms. ecological implications. Certain bird species 0.98), suggesting that T. radicans embryos eat poison ivy drupes (Martin et al., 1951; were not under physiologically enforced Discussion Senchina, 2008). It follows that T. radicans embryo dormancy. Similarly, treatments drupe dispersal and seedling germination that removed the exocarp (water or bleach, Poison ivy (T. radicans) is well known for may be adapted to passage through avian Fig. 4A–B) and partial mesocarp removal its ability to cause the distinguishing delayed intestinal tracts, where the combination of (sandpaper or mechanical scarification with or contact dermatitis symptoms. Although the mechanical grinding in the gizzard coupled without bleach, Figs. 4C and 4D, respectively) chemical composition of T. radicans uru- with strong stomach acid may promote endo- did not significantly increase seedling germi- shiols is documented, there are vast gaps in carp pitting, thus priming the drupe for envi- nation (Figs. 1 and 2). In contrast, in addition knowledge about the physiology and metab- ronmental moisture to reach the embryo. A to the removal of exocarp and most mesocarp olism of urushiol in T. radicans and other study examining the effects of T. radicans tissue, SA treatments also resulted in pitting of closely related plant species. To date, all drupe herbivory by squirrels and Ruffed the outer endocarp (Fig. 4E) that was in turn studies used T. radicans tissues harvested Grouse demonstrated that T. radicans drupes associated with significantly increased seed- from unmanaged plants growing in natural isolated from the feces of Ruffed Grouse ling germination frequencies (Figs. 1 and 2). environments, in which important genetic showed seedling germination rates that were Endocarp pitting specifically removed the and environmental parameters were not con- not significantly different from drupes that impermeable exterior brachysclereid and ad- trolled. Based on that experimental approach, were not eaten by the birds (Penner et al., jacent osteosclereid cell layers (Fig. 4E), re- it is difficult to establish whether observed 1999). With that said, the mean germination sponsible for keeping water from passing differences in urushiol accumulation levels frequency of control water-treated drupes in through the endocarp. The thick and more and/or composition were the result of any of the present study ranged from 1% to 3% porous internal macrosclereid cell layer a wide range of parameters including plant (Fig. 1), whereas the untreated drupe germi- remained mostly intact at the pitted endocarp tissue type, age, genotype, environmental (e.g., nation frequency in Penner et al. (1999) ranged regions (Fig. 4F). Over most of the endocarp, light quality/quantity, moisture, and mineral from 12.5% to 16%, indicating the drupes were the macrosclereid cell layer was relatively nutrients) and biotic or abiotic stress. Thus, substantially more prone to germinate before thick. The notable exception was the micropy- more exacting studies of T. radicans urushiol being eaten by the Ruffed Grouse. Poison ivy lar region where the macrosclereid layer was physiology will necessitate controlling as shows a high degree of anatomical poly- much thinner and from which the germinating many of the aforementioned experimental morphism in different geographical locations

1528 HORTSCIENCE VOL. 48(12) DECEMBER 2013 and variation of urushiol concentrations within one plant. Phytochemistry 19:799–802. Craig, J.C., C.W. Waller, S. Billets, and M.A. Elsohly. 1978. New GLC analysis of urushiol congeners in different plant parts of poison ivy, Toxicodendron radicans. J. Pharmacol. Sci. 67:483–485. Gillis, W.T. 1971. The systematics and ecology of poison-ivy and the poison-oaks (Toxicoden- dron, Anacardiaceae). Rhodora 73:72–159, 161–237, 370–443, 465–540. Johnson, R.A., H. Baer, C.H. Kirkpatrick, C.R. Dawson, and R.G. Khurana. 1972. Comparison of the contact allergenicity of the four penta- decylcatechols derived from poison ivy uru- shiol in human subjects. J. Allergy Clin. Immunol. 49:27–35. Kurtz, A.P. and C.R. Dawson. 1971. Synthesis of compounds structurally related to poison ivy urushiol. 3. 3-n-pentadecylcatechol and 3-n- alkylcatechols of varying side-chain length. J. Med. Chem. 14:729–732. Li, X., J.M. Baskin, and C.C. Baskin. 1999a. Anatomy of two mechanisms of breaking physical dormancy by experimental treat- ments in seeds of two North American Rhus species (Anacardiaceae). Amer. J. Bot. 86: 1505–1511. Li, X., J.M. Baskin, and C.C. Baskin. 1999b. Fig. 4. Effects of treatments on subsequent poison ivy drupe anatomy. (A) Unscarified with water Physiological dormancy and germination re- treatment; (B) unscarified with bleach treatment; (C) sandpaper scarified with bleach treatment; quirements of seeds of several North American (D) sand/rock scarified with water treatment; (E) sand/rock scarified with sulphuric acid (SA) and then Rhus species (Anacardiaceae). Seed Sci. Res. bleach treatment; and (F) medial section of endocarp released from germinated seedling in response 9:237–245. to sand/rock scarification with SA and bleach treatment. Drupe features indicated by arrows Li, X., J.M. Baskin, and C.C. Baskin. 1999c. Seed and/or letters: ex = exocarp; m = mesocarp; r = resin duct; en = endocarp; b = brachysclereids; morphology and physical dormancy of several o = osteosclereids; ms = macrosclereid; p = pitting; sc = seedcoat; and asterisk = carpellary North American Rhus species (Anacardiaceae). microphylar region. Bars equal 1 mm. Seed Sci. Res. 9:247–258. Martin, A.C., H.S. Zim, and A.L. Nelson. 1951. American wildlife & plants. Dover Publica- (Gillis, 1971), so the degree to which dif- An appreciable portion of T. radicans tions, Inc., New York, NY. ferences in environmental and/or genetic drupes were effectively surface-sterilized, Mohan, J.E., L.H. Ziska, W.H. Schlesinger, R.B. factors contributed to the different patterns resulting in axenic seedlings. In the RoaCo-1 Thomas, R.C. Sicher, K. George, and J.S. of T. radicans embryo dormancy in the present T. radicans ecotype, 25% of the drupes were Clark. 2006. Biomass and toxicity responses study and Penner et al. (1999) are currently initially scored as sterile and needed to be of poison ivy (Toxicodendron radicans)to unknown. On the other hand, T. radicans drupes physically removed from adjacent fungal- elevated atmospheric CO2. Proc. Natl. Acad. eaten by squirrels were non-viable because contaminated drupes before fungal hyphal Sci. USA 103:9086–9089. squirrels chew through the endocarp, to re- growth spread to contaminate the sterile Murashige, T. and F. Skoog. 1962. A revised move and then eat the embryo, whereas drupes. Thus, the visual identification of medium for rapid growth and bio assays with birds swallow the drupes intact and do not sterile drupes at low plating density com- tobacco tissue cultures. Physiol. Plant. 15:473– 497. separate the embryo from the endocarp. Based bined with their timely removal were essen- Penner, R., G.E.E. Moodie, and R.J. Staniforth. on the particulars of that study, birds are at tial cultural practices for producing axenic 1999. The dispersal of fruits and seeds of least efficient vectors for T. radicans drupe T. radicans seedlings. Because T. radicans Poison-ivy, Toxicodendron radicans,byRuffed dispersal. is a dioecious species (Gillis, 1971), harvest- Grouse, Bonasa umbellus, and squirrels, Microbial contamination of T. radicans ing drupes from a single female plant ensures Tamiasciurus hudsonicus and Sciurus caroli- drupes was a chief impediment for attaining that the drupes are all genetically related as nensis.Can.FieldNat.113:616–620. axenic T. radicans seedlings. T. radicans half-sibs. Thus, the ability to culture axenic Senchina, D.S. 2008. Fungal and animal associates drupes treated only with sterile water showed T. radicans half-sib seedlings provides un- of Toxicodendron spp. (Anacardiaceae) in extensive bacterial contamination as evi- precedented opportunities for well-controlled North America. Perspect. Plant Ecol. Evol. Syst. 10:197–216. denced by copious slimy microbial growth. studies in urushiol metabolism and physiol- Symes, W.F. and C.R. Dawson. 1953. Separation Although 50% bleach treatment was effec- ogy. Given the demonstrated effects of in- and structural determination of the olefinic tive at greatly reducing bacterial growth, it creased atmospheric CO2 levels resulting in components of poison ivy urushiol, cardanol was not adequate to efficiently sterilize fun- increased poison ivy growth and the produc- and cardol. Nature 171:841–842. gal spores or hyphae. Likewise, combining tion of more allergenic urushiol congeners Symes, W.F. and C.R. Dawson. 1954. Poison ivy 50% bleach with concentrated SA was also (Mohan et al., 2006; Ziska et al., 2007), detailed ‘urushiol.’ J. Amer. Chem. Soc. 76:2959–2963. inefficient at surface sterilization of either knowledge of urushiol metabolism and phys- Talley, S.M., R.O. Lawton, and W.N. Setzer. 1996. sandpaper-scarified or mechanically scari- iology is required to develop novel T. radi- Host preferences of Rhus radicans (Anacardia- fied drupes. The recalcitrance of the drupe- cans control measures. ceae) in a southern deciduous harwood forest. Ecology (Wash. D. C.) 77:1271–1276. associated fungi to 30 min of concentrated Literature Cited Ziska, L.H., R.C. Sicher, K. George, and J.E. SA treatment is particularly notable, sug- Mohan. 2007. Rising atmospheric carbon di- gesting that the fungi might be present as Baer, H., M. Hooton, H. Fales, A. Wu, and F. oxide and potential impacts on the growth mesocarp endophytes rather than mere sur- Schaub. 1980. Catecholic and other constitu- and toxicity of poison ivy (Toxicodendron face saprophytes. ents of the leaves of Toxicodendron radicans radicans). Weed Sci. 55:288–292.

HORTSCIENCE VOL. 48(12) DECEMBER 2013 1529