Journal of Chemical Ecology. VoL 21, No. 9, 1995

SECONDARY CHEMISTRY OF HYBRID AND PARENTAL WILLOWS: PHENOLIC GLYCOSIDES AND CONDENSED TANNINS IN Salix sericea, S. eriocephala, AND THEIR HYBRIDS

C O L I N M. O R I A N S j'* and R O B E R T

S, F R I T Z -~

IDepartment t)J"Biology. Williams College Willktmstown, Massachusetts 01267 "-Department of Bioh~gy, Vassar College Poughkeepsie. Nen" York /2601

(Received November 10, 1994; accepted April 8, 1995) Abstraet--&dix sericea and ,7. eriocephala differ markedly in secondary chemistry. S. sericea produces phenolic glycosides, salicortin and 2'-cinnamoylsalicortin, and low concentrations or condensed tannin, In contrast, S. eriovephala produces no phenolic glycosides, but high concentrations o1"condensed tannins. Hybrid chemistry is intermediate |or both types of chemicals, suggesting predominantly additive inheritance of these two defensive chemical systems from the parental species. However, there is extensive variation among hybrids. This variation may be due to genetic variation among parental genotypes, which genes were passed on, or to subsequent back-crossing. The differences in chemistry are likely to exert a strong effect on the relative susceptibility of hybrid and parental willows to herbivores. Key Words--Salicaceae. willows, hybrids, hybridization, phenolic glycosides, salicorlin, 2'-cinnamoylsalicorlin, condensed tannins.

INTRODUCTION H y b r i d i z a t i o n a m o n g plants is very c o m m o n (Barton and H e w i t t , 1985; R i e s e berg a n d B r u n s f e l d , 1992; R e i s e b e r g a n d E l l s t r a n d , 1993). T h e ubiquity o f h y b r i d s has i m p o r t a n t i m p l i c a t i o n s to plant s p e c i a t i o n and c o u l d affect the e v o lution o f plant p h y t o p h a g e s . In fact, the study o f h e r b i v o r e s in plant h y b r i d *To whom correspondence should be addressed. Present Address: Department of Biology Tufts University Medford, MA 02155. 1245 2751 3028/95109(~-1245507.50/0!~:~1995PlenumPublishingCorpt*r'alion

1246

ORIANS AND FRITZ

zones has recently gained attention (Whitham, 1989; Boecklen and Spellenberg, 1990; Whitham et al., 1991, 1994; Aguilar and Boecklen, 1992; Floate and Whitman, 1993; Paige and Capman, 1993; Fritz et al., 1994). While some have argued that hybrids typically have higher densities of herbivores and thus may serve as ecological and evolutionary sinks (sensu Whitham 1989), others have argued that the relative abundance of herbivores and hybrids and parental taxa is more variable; densities may be greater, less, or equal to the parental species (Boecklen and Spellenberg, 1980; Aguilar and Boecklen, 1992; Fritz et al., 1994), The different patterns of herbivore abundance could be due to the specifics of the plant system; Boecklen and colleagues study oaks. Whitham and colleagues study cottonwoods, and we have been studying willows. The morphology and chemistry of hybrids will differ among systems (Meier et al., 1989: O'Donoughue et al., 1990; Levy and Milo, 1991; Reiseberg and Ellstrand, 1993). However, differences in abundance could merely be due to variation among herbivores in their response to plant traits (Rowell-Rahier, 1984; Tahvanainen et al., 1985a; Marquis, t990). The effects of hybridization on plant morphology are well researched (Whitham, 1989; Boecklen and Spellenberg, 1990; O'Donoughue et al., 1990; Levy and Milo, 1991; Reiseberg and Ellstrand, 1993), but the effects of hybridization on foliar chemistry are less understood (Rabotyagov and Akimov, 1987: Huesing et al., 1989: Meier et al_, 1989; Reiseberg and Ellstrand, 1993). Willows and other saticaceous plant species produce two main secondary chemicals: phenolic glycosides and condensed tannins (Julkunen-Tiitto, 1986, 1989: Lindroth et al., 1987), In this system, S. sericea produces high concentrations of phenolic glycosides, especially salicortin (Nichols-Orians et al,, 1992, 1993) and phenolic glycosides are well known to influence the susceptibility of plants to both insect and mammalian herbivores (Rowell-Rahier, 1984; Tahvanainen et al., 1985a,b: Basey et al., 1988; Lindroth and Peterson, 1988; Lindroth et al., 1988; Clausen et al., 1989; Denno et al., 1990; Kolehmainen et al., 1994). Preliminary results indicated that S. sericea produces low concentrations of condensed tannins, while S. epTocephala produces high concentrations of condensed tannins, but no phenolic glycosides. Since condensed tannins can be bioactive as well, especially against microbes (Zucker, 1983; Schultz, 1989), the patterns of inheritance of these two chemicals could have strong effects on the relative resistance of hybrids to phytopbages. Here we report on the concentrations of condensed tannin and phenolic glycosides in Salix sericea, Salix eriocephala, and hybrids of the two species. To avoid confounding environmental effects (Bryant et al., 1987: Price et al., 1989; Julkunen-Tiitto et al., 1993), we grew the hybrids and the two parental taxa in a common garden.

t247

S E C O N D A R Y CHEMISTRY OF WILLOWS

METHODS

AND MATERIALS

Plants. Both Salix sericea Marshall and Salix eriocephala Michx. are abundant in the northeastern United States and eastern Canada (Argus, 1986). They produce leaves continuously from May until September at our field site near Milford, New York, and adult plants reach a height of 3-4 m. Hybrid and parental plants are abundant within our field site. In March 1993 we took leafless budsticks (cuttings) from seven S. eriocephala, eight hybrid, and eight S, sericea field plants. These plants were far apart from one another and therefore likely represent different genotypes. Cuttings were dipped in Rootone and planted in a soiless medium until roots formed. In May the plants were transplanted into pots containing soil, peat moss, and vermiculite (3 : 1 : I). For this experiment one cutting from each of the field plants was used. Plants were replaced in a common garden on the Williams College campus, and fertilized weekly with 120 ml of 3.2 g/liter 2020-20 Agway complete fertilizer. On July 20, leaves were collected for chemical analyses, Leaf Material. Leaf developmental age can have pronounced effects on leaf chemistry (unpublished data). Therefore it is important to compare leaves of similar age. The results reported here are based on the chemistry of the first fully mature leaf. In order to get enough leaf material for both tannin and phenolic glycoside analyses, we collected 10 different leaves (the first fully mature leaf on 10 different shoots). Leaves were cut at the petiole, placed on ice, transported to the laboratory and vacuum-dried. Vacuum-drying of leaves has been shown to prevent the loss of both condensed tannins and phenolic glycosides (Orians, 1995). Chemical Analysis. The phenolic glycosides were assayed using standard techniques (Nichols-Orians et al,, 1992, Orians, 1995). Briefly, leaf powder (30 + 3 mg) was extracted in cold MeOH (10 mg leaf powder/ml MeOH) with sonication for 10 rain. Cold water was constantly flushed through the sonicator to prevent the heating of samples. We centrifuged and filtered (0.45 #m filter) each sample before placing extracts in crimp-top vials. Extracts were kept in the freezer until analysis (48 hr or less). We quantified the concentration of glycosides with an HPLC (Hewlett-Packard) and a UV detector set at 274 nm. A reverse-phase Nova-Pak C~8 (4 #m) column (Waters) and a gradient system of distilled water and MeOH was used, We injected samples through an LS3200 automated sampler (SGE). 1,3-Dimethoxybenzene was used as the internal standard. Standard curves were determined for salicortin and 2'-cinnamoylsalicortin. The condensed tannins also were extracted using standard techniques (Orians, 1995). Briefly, approximately 300 mg leaf powder was washed with ether

1248

ORIANSAND FRITZ

and extracted with 70% acetone for 3 hr in a 40°C water bath. Acetone was removed under reduced pressure, and all extracts were diluted to 5 ml with distilled water. The butanoI/HCI method was used to quantify tannin concentrations (Hagerman and Butler, 1989). The concentration of the condensed tannin (milligrams per gram dry leaf) was determined using purified tannins collected from the two willow species and the hybrids. For reference we have included results using quebracho tannin as a standard.

A so t • Salicortin ~" 7ot ~ 2'-c:|nnam

°! '°I ¢J

~o o

NONE

S. eriocephala

hybrid

S+ sericea

S, eriocephala

hybrid

S+ sericea

z zz

zO ~er" i¢1 I-" tuO clZ

zO OCJ o

TAXON

FtG. 1, The concentration (milligrams per gram dry leaf weight) of (A) phenolic glycosides and (B) condensed tannins in parental and hybrid willows. Arrows represent midpoint concentrations,

1249

SECONDARY CHEMISTRY OF: WILLOWS RESULTS A N D DISCUSSION

The concentrations of condensed tannins and phenolic glycosides differ among the three taxa (ANOVA, P < 0.03) (Figure 1). S. sericea contains high concentrations of salicortin and 2'-cinnamoylsalicortin and low concentrations of condensed tannins. Conversely, S. eriocephala produces high concentrations of condensed tannins but no phenolic glycosides. The concentrations of these two types of chemicals are intermediate in the hybrids. We pertbrmed contrasts to determine whether the concentrations in hybrids are statistically different from the parental midpoint. We found no significant difference for condensed tannins (P -> 0.06), salicortin (P = 0.51) or 2'-cinnamoylsalicortin (P = 0.39). The intermediate concentrations of these chemicals suggests additive inheritance [a result consistent with other studies (Meier et al., 1989; Reiseberg and Ellstrand, 1993] or balanced ambidirectional dominance. Deviation from the midpoint would have indicated some directional dominance. The concentrations of both condensed tannins and phenolic glycosides among hybrids are quite variable. In general, the coefficients of variation resemble S. eriocephala for tannin concentration and S. sericea for phenolic glycoside concentration (Table 1). Thus, overall, hybrids are more variable and the lack of significant differences, on average, from the mid-parent reflects the plants we chose to study. Condensed tannin concentrations in hybrids range from 78 to 171 mg/g dry leaf weight, and salicortin and 2'-cinnamoylsalicortin concentrations range from 10 to 70 and from 0.75 to 5.6 mg/g dry leaf weight, respectively. The variation among hybrids may reflect parental variation in chemistry.

TABLE | . COEFFICIENT OF VARIATIONS FOR CONCENTRATIONS OF PHENOLIC GLYCOSIDES AND CONDENSED TANNINS IN

S. sericea, HYBRID

AND S. eriocephala LEAVES Taxon Chemical Phenolic glycosides Salicortin 2'-Cinnamoylsalicortin Condensed tannin Absoluteconcentration %QTE"

S. sericea

Hybrid

S. eriocephala

46.1 59.5

42.5 39.9

NA NA

3.5 2.0

38.6 32.1

28.9 30.4

* %QTE = % quebracho tannin equivalents.

1250

ORIANS AND FRrrZ

A hybrid may produce a low concentration of 2'-cinnamoylsalicortin if its S. sericea mother does. We know that phenolic glycoside concentrations vary extensively within S. sericea (Nichols-Orians et al., 1993), and have found that S. sericea female #61 and its S. sericea progeny contain low concentrations of 2'-cinnamoylsalicortin (unpublished data). We are currently testing the correlation between parental and offspring chemistry for both intraspecific and interspecific crosses. There could also be between-progeny variation because a parent passes on half its genes to each offspring. Finally, variation among hybrids could be due to whether the hybrid is F~, F~_, or back-crossed to one of the parental species. We do not know the status of these hybrids, but are currently collaborating with Stephen Brunsfeld to evaluate the effects of back-crossing on the chemistry. Interestingly, we found an inverse correlation between tannin and phenolic glycoside concentration among the hybrids (Figure 2A and B). (No negative correlation was found within S. sericea.) This negative phenotypic correlation indicates that it may be impossible for hybrids to produce high concentrations of both phenolic glycosides and condensed tannins (sensu Berenbaum et al., 1986). Future studies will evaluate the basis of this trade-off and its effect on hybrid-herbivore interactions. Differences in chemistry can strongly affect the resistance of salicaceous plants to herbivores (Rowell-Rahier, 1984; Tahvanainen et al., 1985a; Basey et al., 1988; Lindroth and Peterson, 1988; Lindroth et al., 1988: Ctausen et al,, 1989; Denno et al., 1990; Kolehmainen et al., 1994). For example, Tahvanainen et al. (1985a) found that herbivory by several willow-feeding beetles varied as a function of the phenolic glycoside content of different willow species. They found that the response of each beetle species depended upon its adaptations to phenolic glycosides. Previously we reported (Fritz et al,, 1994) that many herbivore species differ in their response to S. eriocephala, S. sericea, and S.e. × S.s. hybrids. Depending on the herbivore, hybrids were less resistant, intermediate in resistance, or as resistant as one of the parental taxa. Since we know phenolic glycosides and tannins differentially affect herbivores (Rowell-Rahier, 1984; Tahvanainen et al., 1985a; Schultz, 1989; Denno et al., 1990), plant chemistry may be one factor that explains the varied herbivore responses. There is, however, no a priori reason to expect hybrids to be more or less resistant than either of the parental taxa. The apparent additive inheritance of these secondary chemicals in hybrids has further implications for the resistance of hybrids to herbivores. Whitham (1989) suggests that back-crossing might lead to the further breakdown of resistance traits of backcross plants compared to parents or F~s. However, if a resistance trait were additively inherited, back-crossing to the parent with the higher concentration should lead to the recovery of resistance traits of the parent. This

SECONDARY CHEMISTRY OF WILLOWS

1251

A

5° L-1 ~ "

y = 98.184 - 0.50x R2 = 0"87

v

z Io .-I

u)

50

100

150

200

o y v Z 1--

=

R2

6.51

-

0.03x

= 0,62

6'

o B .J

03 J >o

=E

4 •

z.

< z

z

0

100

150

CONDENSED TANNIN

200

(MG/G)

FIG, 2. The relationship between (A) condensed tannin and salicortin concentration, and (B) condensed tannin and 2'-cinnamoylsalicortin concentration in hybrid leaves.

in turn should result in a parallel herbivore response rather than further breakdown. We are currently studying both the inheritance of chemical variation and resistance in backcross plants to test this hypothesis. Acknowledgments--We thank Len and Ellie Sosnowski and the New York State Department of Environmental Conservation for permitting us to collect plants from their property. Rebecca Barnes provided invaluable assistance• We are especially indebted to Dr. Thomas Clausen for providing us with purified tannins and to Dr. Dan Lynch for letting C. Orians u ~ his HPLC. The Department of Biology and the Bronfman Science Center at Williams College provided equipment support. This research was supported by the National Science Foundation (grant DEB 92-07363).

1252

ORIANS AND F r n ' z REFERENCES

AGUILAR,J.M., and BOECKLEN,W.J. 1992. Patterns of herbivory in the Quercus grisea x Quercus gambelii species complex. Oikos 64:498-504, ARC;US. G.W. 1986. The genus Salix lSalicaceae) in the southeastern United States. Syst. Bot. Monogr. 9:1-170. BARTON, N.H., and HEWFr, G.M. 1985. Analysis of hybrid zones. Annu. Rev. EcoL System. 16:113-148. BASE",', J.M., JENKINS, S.H., and BUSHER, P.E. 1988. Optimal central-place foraging by beavers: Tree size selection in relation to defensive chemicals of quaking aspen. Oecotogia 76:278 282. BErENBAU~',t, M.R., ZANGERt., A.R., and NI'rAo. J.K. 1986. Constraints on chemical coevolution: Wild parsnips and the parsnip webwom~. Evohaim~ 40:1215-1228. BOECKLEN, W.J., and SPELLENBER~, R. 1990. Structure of herbivore communities in two oak (Quercus spp.) hybrid zones. Oe('nh~gia 85:92-100. BRYANT, J.P., CLaUSEN, T.P., REICHArDT, P.B., MCCARrttY, M.C., and WERnER, R.A. 1987. Effect of nitrogen lerfilization upon the secondary chemistry and nutritional value of quaking aspen (Populus tremuloides Michx.) leaves for the large aspen torlrix I Choristoneura cot~flictana (Walker)]. Oecolngia 73:513 517. CLAUSEN. T.P., REICHarD'r, P.B.. BRYANT,J.P., WERNEr, R.A., Pt~ST. K., and FRISBV. K. 1989. Chemical model lot short-tern1 induction in quacking aspen (Populu.~, tremuloides) foliage against herbivores. J. Chem. Ecol. 15:2335-2346. DENNO, R.F., LARSSON,S., and Ot.MSrE,',D, KL. 1991). Role of enemy-free space and planl quality in host-plant selection by willow beetles. Ecology 71: 124-137. FLO.,~'rE, KD., and WHITMAN,T.G. 1993. The "hybrid bridge '~ hypothesis: Host shifting via phmI hybrid swarms. Am. Nat. 141:651-662. FRITZ, R.S.. NICHOLS-ORIANS,C.M., and Brt~NSEEhI~. S.J. 1994. lnterspecific hybridization of plants and resistance to herbivores: Hypotheses, genetics, and variable responses in a diverse herbivore community. Oecologiu 97: 106-117, HAGERMAN,A.E., and BUTLEr. L.G. 1989, Choosing appropriate methods and standards lot assaying tannin. J. Chem. Ecol. 15:1795-1810. HUESING. J., JONES. D.. DEVI-RNA, J., MYERS. J., COlA.INS. G., SI'VERS~)N, R., and SISSoN. v. 1989. Biochemical investigations of antibiosis material in leaf exudate of wild Nicotiamz species and interspecific hybrids. J. Chem. Ecol, 15:1203-1217. JULKUNEN-TtITro, R. 1986. A chemotaxonomic survey of pheno[ics in leaves tff northern Salicaceac species. PhytochemL~'try 25:663-667. JULKUNEN-THTTO, R. 1989. Phenolic constituents of Salix: A chemolaxonomic survey of further Finnish species. Phytochemistry 28:2115-2125. JULKtmEN-TIITTO, R., TAHVANaINEN,J., and SJLVOLA,J. 1993. Increased CO, and nutrient status changes affect phytomass and the production of plant defensive secondary chemicals in SalLr myrsinifi)lia (Salisb.). Oecoh~gia 95:495-498. KOLEHMAINEN, J., ROININEN,H., JULKUNEN-TH'I-FO,R., and TAHVANAINEN,J. 1994. Importance of phenolic glucosides in host selection of shoot galling sawfty, Euura atne14nae, on Salix pentandra. J. Chem. Ecol. 20:2455-2466. LEvy. A., and MILO, J. 1991. Inheritance of morphological and chemical characters in interspecific hybrids between Papaver bracteatum and Papuver pseudo-orientale. Theor. Appl. Genet. 81:537-540. LINDROTH, R.L., and PETERSON, S.S. 1988. Effects of plant phenols on perlormance of southern arrnyworm larvae. Oecologia 75:185-189.

SECONDARY CHEMISTRY OF WILLOWS

1253

LINDRO.TH, R.L., HSIA, M.T.S,, and SCR~Bt-R, J.M. 1987. Chardcterization of phenolic glycosides from quaking aspen. Biochem. Svst. Ecol. 15:677-680. LINDrOTH, R.L., SCRIBEr, J.M,. and HSlA, M,T.S. 1988. Chemical ecology of the tiger swallowtaih Mediation of host use by phenolic glycosides. Ec,logy 69:814-822. MARQUIS, R.J. 1990, Genotypic variation in leaf damage in Piper arieianum (Piperaceae) by a multispecies assemblage of herbivores. Evolution 44: 104-t20. MEIER, B., BETTSCHARI, A., SFIAO, Y., and LAUTENSCHLAGER, E. 1989. Einsatz der modemen HPLC fur chemotaxonomische Untersuchungen morphologisch schwer zu differenzierender Sali.r-Hybriden. Pkmm Meal. 55:2t3-214. N~CHOt.S-ORIANS, C,M., CLAUSEN, T.P., FRITZ, R.S., REtCHArDT, P.B., and Wt~, J. 1992. A new phenolic glycoside isolated from Salix serieea Marshall. Phytochemistry 31:2180-2181. NICHOLS-OrlANS, C , M . Frrrz. R.S., and Ct_AUSEN, T.P. 1993. The genetic basis lot variation in the concentration of phenolic glycosides in Salix serieea: Clonal variation and sex-based differences. Bh~chem. ,~vst. Ecol. 21:535-542. O'DONot/GHUE, L.S., RAELSON, J.V. and GRANT, W.F. 1990. A morphological study ot+interspecific hybrids in the genus IJ~tus (Fabaceael, Can. J. Bot. 68:803-812. ORmANS, C.M. 1995. Preserving leaves for tannin and phenolic glycoside analyses: A comparison of methods using three willow taxa. J. Chem. Ecol, 21:1235-1243. PaIGt-~, K,N,, and CAPMAN, W.C. 1993. The effects of host-plant genotype, hybridization, and environment on gall-aphid attack and survival in cottonwood: The importance of genetic studies and the utility of RFLPS. Evcdution 47:36-45. PRICE, P.W., WARING, G.L., JULKLINEN-TItTTO, R., TAHVANAINEN,J., MOONEY, H.A., and CRAIG, T.P. 1989. Carbon-nutrient balance hypothesis in within-species pbytochemical variation of Salix htsiolepis. J. Chem. Ecol. 15:1117-1131. R,~B(~'r'~'~GOV, V.D. and AKIMOV, Y.A. 1987. Inheritance of essential oil content and composition in interspecific hybridization of lavender. Soy. Genet. 22:1163-1172. REISElaERG, L.H,, and BRUNSEEtD. S.J. 1992, Molecular evidence and plant intmgression, pp. 151176, in P.S. Soltis, D.E. Sollis, and J.D. Doyle (eds.L Molecular Systematics of Plants. Chapman and Hail. New York. RIESEBERG, L.H., and ELLSTRAND, N.C. 1993, What can molecular and morphological markers tell us about plant hybridization, Crit. Rev. Plant Sci. 12:213-241. ROWELL-RAHU-~R, M, 1984, The presence or absence of phenolglycosides in Salix (Salicaceae) leaves and the level of dietary specialization of some of their herbivorous insects. Oecologia 62:2630. SCHULTZ, J.C. 1989, Tannin-insect interactions, pp. 417-433, in R.W. Hemingway and J.J, Karchesy (eds,), Chemistry and Significance of Condensed Tannins. Plenum Press, New York. TAHVANAINEN, J., JULKUNEN-TIII-rO. R., and KETTUNEN, J. t985a. Phenolic glycosides govern the food selection pattern of willow feeding leaf beetles. Oeeoh)gia 67:52-56. TAHVANA~NEN, J., HELLE, E., JULKUNEN-TIITTO, R. and LaVOLA, A. 1985b, Phenolic compounds of willow bark as deterrents against feeding by mountain hare. Oecologia 65:319-323. WHrmam, T.G. 1989. Plant hybrid zones as sinks lbr pests. Science 244: 1490-1493. WHm~AM, T.G.. Morrow, P,A., and Poyrs, B.M, 1991. Conservation of hybrids. Science 254:779780. WHITHAm, T.G., Morrow, P . A , and Po'rTs, B.M. 1994. Plant hybrid zones as centers of biodiversity: The herbivore community of two endemic Tasmanian eucalypts. Oeeologia 97:481490. ZUCKER, W.V. 1983. Tannins: Does structure determine function'? An ecological perspective. Am, Nat. 121:335-365.