Autotrophy and heterotrophy in root herniparasites

,‘, ,,/;,

I-

.,

TREE vol. 4, no. 9, September

,I

Autotrophyand Heterotrophyin RootHemiparasites Malcolm C. Press More than 3000 speciesof floweringplants are at least partially parasitic, acquiring water and solutes from the host via haustoriu. More than one third of all parasitic angiosperms - the root hemiparasites possess green leaves and root systems. In these species there are potentially two opportunities for the capture of water and solutes: an autotrophic or a6iotic supply from the external environment, and a heterotrophic or host-derived supply via the haustoria. Most root hemiparasites occur in the Scrophulariaceae, a family also containing autotrophic and holoparasitic plants. Between these two extremes, the root hemiparusites provide an ideal opportunity to investigate the balance 6etween the autotrophic and heterotrophic modesof nutrition in parasitic plants. The tropical hemiparasites within this family are important weeds of cereals and legumes, causing considera6le crop losses, and thus fuelling research into the nutritionul dependency of these plants on their hosts. These studies have led to some exciting new ideas, particularly with respect to the carbon relations of these plants.

trophy are normally considered with respect to carbon acquisition, but here they are used in their strict sense to refer to both inorganic and organic solutes. A fuller understanding of heterotrophy in parasitic angiosperms may enable us to develop improved control strategies for those parasites that are important weeds of cereals and legumes. The most important of these is Striga, which can reduce grain yields of sorghum, maize and millet to virtually nothing3v4.

Autotrophy:growthand soluteacquisition Most of the root hemiparasitic Scrophulariaceae are obligate parasites, but many genera can survive to the seedling stage in the absence of a hosts, and a few (the facultative parasites) can flower and set viable seed (e.g. Rhinanthus6, Euphrasia7f8, Orthocarpusg, Bellardiag and Paren tucellia9). In nature, however, unattached mature facultative parasites are rare, and attachment to a host greatly stimulates their growth&g. The extent of growth stimulation depends on the host species”9. For Parasitism has arisen at least example, in pot experiments with eight different times in 17 plant Orthocarpus purpurascens9, growth families’J (see Table I), and parachanges following attachment sitic plants are found from polar ranged from nothing with Trifolium regions to the equator. All parasitic repens to more than a threefold plants possess haustoria, which increase with Spergula arvensis may be located either above (Fig. I). ground (stem parasites) or below Since some of these plants can ground (root parasites). Stem and complete at least part of their life root parasites may be further subcycle in the absence of a host, they divided according to the presence must be capable of autotrophic solor absence of chlorophyll, and are ute acquisition. Axenic (i.e. pure) termed hemi- (or semi-) and holoparasites respectively. culture of both obligate (Striga and AlectraIl and facultative (OrthoMost root hemiparasites occur in carpus)” parasites shows that these the Scrophulariaceae. The presence of a root system and green genera can complete their life leaves as well as haustoria suggests cycle on simple inorganic media with the addition of sugars. Howthat root hemiparasites may receive and incorporate water and ever, plants treated in this way solutes from both their abiotic have a weak and spindly habitlo, (autotrophic) and biotic (heteroand their performance is poor comtrophic) environments. The term pared to that of plants growing in autotrophy and its antonym heteronatural conditions. It is difficult to assess the extent of autotrophy from these experiments, because Malcolm Press is in the Sfriga Research Group, the concentrations of solutes in the Biology Dept, Darwin Building, University College, ‘minimal’ and ‘basal’ media used Cower Street, London WCIE 6BT, UK. (From September 1989: Dept of Environmental Biology, are one to three orders of magniThe University, Manchester M I3 9PL, UK.) tude greater than would be en258

Ii2: 989

1989

countered in the soil solution land in host sap). In experiments with Striga12 (Fig. 21, growth was greatly stimulated by the addition of an inorganic nitrogen mixture containing 20 mM NH4+ and 40 mM N03- , to an N-free medium. Growth stimulation thus indicates some ability to take up and assimilate inorganic nitrogen. The in vitro studies with Strjga also demonstrate growth stimulation as a result of organic nitrogen (glutamine) supply (Fig. 21, but at more realistic concentrations, suggesting that this genus is more suited to a heterotrophic nitrogen source. In conclusion, in vitro culture experiments have given little idea of the functional capacity of the root systems of these plants in the soil. There are fewer reports of the nutrition of unattached facultative hemiparasites in soil, but metabolic studies of such plants also demonstrate that they can take up and assimilate nitrate. The first enzyme involved in the assimilation of nitrate (nitrate reductase) is substrate-inducible; its activities reflect the supply of nitrate to the plant. However, it is difficult to determine whether nitrate reductase activities measured in parasites reflect uptake of nitrate from the soil by the parasite or supply of nitrate via host sap. Experiments involving nitrate fertilization of field plots of unattached Euphrasia frigida have demonstrated direct uptake of nitrate from the soil solution (F.I. Rumsey, pers. commun.). However, quantitative information on the autotrophic acquisition of inorganic solutes is sparse. The assimilatory capacity of the leaves of hemiparasites is low; carbon dioxide fixation rates are at the lower end of the range encountered in C3 plants13-15. In addition, rates of respiration are high, and daytime gains of carbon are more than accounted for by night-time losses in some speciesi2,14,15 (Fig. 3). The data in Fig. 3 would seem to preclude the possibility of survival of unattached hemiparasites. However, experiments with Rhinanthus’6 and Odontites 0.F. Hodgson, PhD Thesis, University of Sheffield, 19731 have shown that rates of photosynthesis in unattached plants exceed those measured following attachment. The mech-

TREE vol. 4, no. 9, September

1989

Table I. Families containing parasitic genera and some of their characteristics’-2 Family Scrophulariaceaec

Chlorophyllous

Location

+

Scrophulariaceaec

of haustoria=

Number of genera/speciesb

B

201500

B

6160

Orobanchaceae

-

B

14/160

Lennoaceae

-

B

316

A

l/180 4l20 l/IO 3/l 0

Convolvulaceaec Olacaceae Myzodendraceae Eremolepidaceae

_d +

:: A

+ +

Krameriaceae Loranthaceae Viscaceae Santalaceae

+ + + i-1 + i-1

LauraceaeC Rafflesiaceae

(+I -d _

II30 9150

Hydnoraceae

_

2/l 0

Balanophoraceae

-

18/100

Cynomoriaceae

-

l/2

Podocarpaceae

+

l/l

B A (B) &-?B

l/20 50/I 000 J/500 301250

aA: haustoria attached to above-ground parts of host (stem parasites). B: haustoria parasites). Parentheses indicate occurrences in a small minority of species bNumbers are approximate in some cases CFamily not entirely parasitic, dominated by free-living genera dChlorophyll present at low concentrations only, species considered to be holoparasitic

anism for this is unclear, but the implications are important with respect to the cost to the plant of investment in photosynthetic machinery. It appears, therefore, that both rates of carbon incorporation by the leaves and inorganic solute uptake by the roots in many of the hemiparasites are sufficient to sustain only poor growth, and that the plants are reliant on additional supplies of these resources from the host. Limitationsto autotrophy The most extreme forms of root parasitism are found in the Rafflesiaceaer. Evolution has stripped RaHesia of all irrelevant organs, leaving only a myceliumlike haustorial system for nutrition and a flower for reproduction’. To a lesser extent, the root hemiparasites in the Scrophulariaceae also show alterations in morphological and metabolic characters as a consequence of their specialized mode of nutrition, affecting the balance between autotrophy and heterotrophy in these plants.

Roots and other below-ground organs in free-living plants account for between 40% and 85% of net primary production, depending on the vegetation17. Most communities possess one or more parasitic angiosperms, and the proportion of their biomass allocated to the below-ground parts probably lies below this range’. Among the parasitic angiosperms, perennial

Habits, habitats

and relationships

Annual and perennial herbs, mainly temperate (e.g. Rhinanthus), but range from arctic (e.g. fedicularis) to mediterranean (e.g. Orfhocarpus) to tropical (e.g. Stfiga) Annual and perennial herbs, mainly mediterranean America only, extremely xeromorphic, deserts and sand dunes Cuscuta (dodder), global distribution Tropical and subtropical shrubs and trees Southern beech forests of S. America Shrubs, tropical America, closely related to Viscaceae Xerophytic shrub, America only Mistletoes, mainly tropical/subtropical, but found over wide geographical range Includes sandalwood, but mainly small perennial herbs and shrubs, mostly tropical Cassytha, perennial, mostly coastal Mostly tropical, few mediterranean species (Cytinus), extreme morphological reduction S. and E. Africa, S. America, extreme morphological reduction High-altitude rain forests, extreme morphological reduction Mediterranean herbs, closely related to Balanophoraceae Parasitaxus ustus, parasitic habit questionable attached

to below-ground

parts of host (root

hemiparasites have the most well-developed root systems, for example Bat-Ma and the perennial Pedicularis spp. However, root growth in these plants may be atypicalis. In Pedicularis the root system is diffuse, and growth of one of the axes is often arrested beyond the point of a dichotomy, resulting in one branch being much the smaller, and in some cases its

35

E so

30-

f$

25-

$8 Q6 “8

20-

8$

15-

Spergula arvensis Hypochoerisglabra Erodium botris Lolium multiflorum

Festuca megalura

Fig. I. Growth, measured as mean inflorescence length, in the root hemiparasite Orthocarpus purpurascens grown in pots either unattached or parasitic on one of six different host species. Bars are means from measurements of 21 pots each containing one host and one parasite. Redrawn from Atsatt and Strong.

259

TREE vol.4, no. 9, September

1989

30 ‘5 g

25-

5 2 20CD z z

15-

F 3

lo5-

0

0.7

1.4

2.8

5.5

11.0

21.9

Glutamine concentration (mM) Fig. 2. Growth responses of Striga hermonthica in axenic culture to additions of inorganic nitrogen and glutamine. Black bars: without inorganic nitrogen. Crey bars: with 60 mM inorganic nitrogen. Drawn from data of Okonkwo’z.

dying back completely. The roots of older plants tend to lack root hairs, except near the parasitic connection. Root hairs are only abundant on the seedlings of this genus’8. In temperate regions, root hemiparasites are often short-lived species of grassland and marshland communities. Here, a major barrier to seedling establishment may be the need to develop a sufficiently large root system to compete with those of established perennial plants19. The tropical genus Striga has a more poorly developed root system than the temperate genera. Root caps and root hairs are lacking20, thus reducing the volume of soil exploited. There is evidence that root cap cells are the main sites for abscisic acid synthesis2’; abscisic acid inhibits root extension, and enhances lateral root formation and the production of root hairs in particular2’. Although root

hairs have little effect on the uptake of mobile ions such as nitrate, they promote uptake where diffusive supply is important, as is the case for ammonium and phosphate ions2’. Foliar tissue analyses of unattached and attached Euphrasia frigida plants growing in poor soils in subarctic Sweden show large increases in the concentration of phosphorus (x2.21 and nitrogen ( x 2.8) following attachment (F.1. Rumsey, pers. commun.). Thus, in the absence of the host, phosphorus and nitrogen may limit productivity as a consequence of poor root hair development. Symbiotic associations in plants are the rule rather than the exception; recent studies have highlighted the importance of mycorrhiza in the nutrition of free-living plants. However, the roots of many if not all hemiparasites in the Scrophulariaceae are non-mycorrhiza122; this may limit

Pedicularis svlvatica Rhinanthus minor

Euphrasia nemorosa

Fig. 3. Carbon dioxide exchange rates measured during the day at light saturation Iphotosynthesis, black bars) and at night (respiration. grey bars) in eight species of temperate root hemiparasites growing in the field. Drawn from data of Press,Graves and Sfewartp4.

inorganic solute uptake. Microbial activity in the rhizosphere (i.e. the zone occupied by roots) and rhizoplane (i.e. the root surface habitat) also plays an important role in plant nutritionzl; the ecology of these zones in hemiparasites may also differ from those of free-living plants. The leaves of hemiparasitic and free-living Scrophulariaceae show distinct anatomical differences23. In (a free-living a leaf of Antirrhinum genus), for example, the palisade mesophyll is four cells deep and well developed; the spongy mesophyll below has clear air spaces around cells of large surface area. The leaves of Striga are less well developed, having ill-defined palisade mesophyll and rather dense spongy mesophyll without the large intercellular spaces seen in Antirrhinum. The effective surface area for the assimilation of inorganic carbon is thus reduced. The structure of Striga chloroplasts is not dissimilar to that of free-living plants, but their numbers are lowerI - about one third of those measured in Antirrhinum23. Consequently, the concentration of chlorophyll in root hemiparasites is at the low end of the range observed in free-living plants13J4. In holoparasitic angiosperms many of the carbon and nitrogen assimilating not enzymes are detectable20r24. There may also be metabolic constraints to autotrophy in the root hemiparasiteG4. The activity of the primary CO,-fixing enzyme, ribulose bisphosphate carboxylase (rubisco), is at the low end of the range found in C3 plants. The activities of some of the enzymes involved in the assimilation of inorganic nitrogen, particularly nitrate reductase and glutamate synthase24-26, are also low. Although nitrate reductase has been measured in species towards the autotrophic end of the parasitic continuum, it is barely detectable in genera such as Striga24. The activities of enzymes involved in the biosynthesis of amino acids are not dissimilar to those of free-living plants24; indeed, those involved in biosynthesis of citrulline, the principal amino acid in Striga, may have higher activities than encountered in free-living plants. There are therefore ecological, morphological and metabolic con-

TREE vol. 4, no. 9, September

1989

straints to autotrophy in the root hemiparasites. Little is known about the control of these characters at the molecular level. For instance it would be of interest to know whether the gene sequences for enzymes that are not detectable in hemi- and holoparasites are absent from these plants, or whether there are factors preventing their expression. Heterotrophy:host-parasite transport The xylem in the haustorium of all parasitic plants is continuous with that of the host. All hostparasite fluxes occur apoplastically (i.e. through the cell walls and intercellular spaces27). In holoparasites that are almost completely heterotrophic, close connections with host phloem are also formed, permitting movement of solutes from host symplasm to parasite apoplasm28. Parasitic angiosperms exhibit a continuum between the photosynthetically competent xylem feeders (e.g. the mistletoe genus Phoradendron) and the almost completely heterotrophic phloem feeders (e.g. Orobanche and Cuscuta). Despite the paucity of phloem elements in the haustoria Scrophulariof hemiparasitic aceae29, transfer of solutes from the phloem cannot be discounted. The flux of water and solutes between host and parasite can be analysed in terms of two factors: the driving force for movement and the resistance to water flow in the host-parasite vascular continuum. The former can be measured either as water potential or as relative water content of host and parasite tissue, and both favour movement from host to parasite30a3’. Water movement is facilitated by the unusual characteristics of the stomata in root hemiparasites27. The leaves have many stomata, which appear to be insensitive to environmental factors that would normally cause stomata of other plants to close. Thus, the hemiparasite continues to lose water at night and during periods of moderate droughtIs. The mechanism by which closure is prevented is unknown, but the guard cells may be insensitive to abscisic acid32, which is also involved in the regulation in freeof stomata1 aperture living plants. Transpiration rates in root hemiparasites exceed those

measured in free-living plants by more than an order of magnituder4, and play an important role in the acquisition of solutes from the host’4f33. Heterotrophy:quantification of fluxes Transfer between host and parasite of radiolabelled compounds (e.g. r4C-C02, r4C-urea, 35S-S042-, 32p-p043-, 45Ca2+) has been demonstrated34, although quantification of these fluxes has not been achieved. It has yet to be determined which, if any, of these solutes confer benefit to the growth of

the attached parasite. Axenic culture studies suggest that the parasites are not dependent on the host for the supply of vitamins and plant hormonesI@-12, but such findings should be treated with caution. The high concentrations of inorganic solutes used in in vitro systems can often lead to the production of secondary metabolites, such as polyamines, which may play a role in growth regulation. More is known about the carbon economy of root hemiparasites. The distribution of naturally occurring 12Cand 13C in host and parasite 261

TREE vol. 4, no. 9, September

(4

Parasite

gen transport compound is asparagine, whereas in Striga hermonthica it is citrulline. The latter has a high N/C quotient (0.5) and is only present in trace quantities in host sap. Thus, there is a flux of carbon associated with the transfer of amino-nitrogen. Mistletoes are capable of autotrophic carbon fixation, and are thought to be independent of host photosynthate33,s7. It has been suggested that the water relations of these plants are such that the heterotrophic supply of nitrogen from the host is optimized33J7. However, the obligate flux of carbon with this nitrogen suggests that the mistletoes may be in receipt of a significant amount of host-derived carbon.

Host

Glu

Man

(w

Host Parasite Ci

Asn (N) N/D/Q/E thers

Fig. 4. (al Proportional pie charts of sugars in the sap of sorghum lhostl and Striga (parasite). Glu, glucose; Fru, fructose; SK, sucrose: Man, mannitol. Total concentrations of sugars are I .27 and 6.69 mol m-3 plant sap for host and parasite respectively. (b) Proportional pie charts of amino acids in the sap of sorghum and Striga. Asn, asparagine; Asp, aspartate; Cln, glutamine; Glu, glutamate; Cit, citrulline; Others, I7 other amino-nitrogen solutes. Total amino-nitrogen concentrations are 28.9 I and 18.43 mol m-3 plant sap for host and parasite respectively. Drawn from data of Mallabum, Press and Stewa@.

has been used to assess carbon fluxes35 (see Box I). In Striga hermonthica, the proportion of hostderived carbon can vary according to the host species, and can range from 8% in S. hermonthica parasitic on a susceptible cultivar of maize to 48% in S. hermonthica parasitic on millet. The form in which organic solutes move from host to parasite has been investigated, and analysis of host and parasite sap shows distinct differences in the nature of compounds transported36 (Fig. 4a). Xylem sap in free-living plants usually contains low concentrations of sugars, and in sorghum infected with Striga hermonthica the three major sugars are glucose, fructose and sucrose. In the parasite, however, the concentration of sap 262

1989

sugars exceeds that of the host by a factor of more than five, and almost 60% of the sugar is in the form of the sugar alcohol mannitol, which is absent from the host sap36 (Fig. 4a). Studies with a number of other have shown that hemiparasites 14C02 fed to the leaves of the parasite is rapidly fixed into mannitol and/or galactitol (J.F. Hodgson, PhD Thesis, University of Sheffield 1973 I. The haustoria of these plants are metabolically very active35, and it has yet to be determined whether all of the sugar alcohols result from parasite photosynthesis or whether they are also synthesized from host photosynthate. Large differences in the amino acid profiles of host and parasite sap have been observed (Fig. 4b336. In sorghum, the major organic nitro-

Towards a cost/benefitanalysisof parasitism in angiosperms What are the advantages of parasitism and why are so few plants parasitic? The morphological and metabolic limitations to autotrophy demonstrate one of the principal benefits of the parasitic lifestyle: make angiosperms parasitic smaller investments in structural features and metabolic pathways. This is apparent both along the parasitic continuum and temporally for individual species, for example ontogenic changes in photosynthetic capacity and root hair fonnation. Access to host-derived water and solutes means that the plant does not have to compete with coexisting species for these resources. This ‘uncoupling’ from the environment may also allow exploration of a greater environmental range. The temperature relations of Striga suggest that this may be the case for this genus? high transpiration rates and the cooling evaporative associated allow the plants to grow in regions of higher temperature than metabolic studies would suggest. The costs of parasitism can be viewed in terms of the problems generated by the heterotrophic mode of nutrition. The possession of a functional haustorium is central to the parasitic life style. The selective pressures favouring the evolution of this organ are unknown (see Refs I,2 and 19). The parasite will receive host sap that may contain primary and secondary metabolites at sub- or super-optimal concentrations for

TREE vol. 4, no. 9, September 1989

growth. The haustorium is metabolically active and may serve to regulate such fluxes. Transpiring plants that do not have symplastic phloem connections with the host may have problems in solute recycling and pH regulation. However, quantitative analyses of sap in the host-parasite association suggests that these are overcome to some extent by transfer between xylem streams2’. Despite the importance of high transpiration rates in solute acquisition, these may predispose the plants to water (and solute) stress in extreme conditions. Hemiparasites may have to invest in mechanisms to ensure against such eventualities24p36. However, restrictions at the community level may be more important than those that operate on the individual’9. The parasite has to establish vascular continuity with a suitable host plant. Within the Scrophulariaceae the seeds of three hemiparasitic genera (Striga, Alectra and Tozzia) require the presence of a stimulus in host root exudate in order to germinate’. Haustorial formation in many (if not all) of the root hemiparasites is also, chemically induced’. Processes governing the selection of competent hosts are poorly understood, and merit further investigation. Experiments with Rhinanthus indicate that high population densities are found in conditions where soil fertility is low and the hosts are less favourable, which suggests that the constraints on populations and individuals are different39. Thus parasitism may require specialization at the level of the population, the whole plant and the cell, and development of cost/benefit analyses will require information from all aspects of biology, from molecules to ecosystems.

Physiol. Plant. 38, 121-125 7 Wilkins, D.A. (1963) Ann. Bot. 27,533-552 8 Yeo, P.F. f 1964) Watsonia 6. l-24 9 At&t, P.R. and Strong, D.R. ( 1970) Evolution 24,278-29 I IO Okonkwo, S.N.C. (1966) Am. /. Bot. 53, 679-687 I I Heineman, R.T. and Atsatt, P.R. ( 1978) 1. Exp. Bot. 29,789-796 I2 Okonkwo, S.N.C. (1966) Am. 1. Bot. 53, 687-693 13 de la Harpe, AC., Visser, J.H. and Grobbelaar, N. (1981) Z. Pflanzenphysiol. 103,265-275 I4 Press, MC., Graves, J.D. and Stewart, G.R. f 1988) 1.Exp. Bot. 39, 1009-1014 I5 Press, M.C., Tuohy, I.M. and Stewart, G.R. (1987) Plant Physiol. 84,814-819 I6 Klaren, C.H. and Janssen, G. ( 19781 Physiol. Plant. 42, 151-155 I7 Fitter, A.H. ( 1987) New Phytol. 106 (SuppI.), 61-77 I8 Piehl, M.A. (1963) Am. 1. Bot. 50,978-985 I9 Watkinson, A.R. and Gibson, C.C. (1988) in P/ant Population Ecology (Davy, A.j., Hutchings, M.1. and Watkinson, A.R., eds), pp, 393-41 I, Blackwell Scientific Publications 20 Musselman, L.l. (1980) Anno. Rev. Phytopathol. 18,463-489 21 Marschner, H. ( 1986) MineralNutrition in Higher P/ants, Academic Press 22 Harley, J.L.and Harley, E.L. ( 1987) New Phytol. 105 fsuppl.), 1-102 23 Tuohy, I., Smith, E.A. and Stewart, G.R. ( I9861 in Biology and Control of Orobanche fter Borg, SJ.. ed.), pp. 86-95, LHNPD Wageningen 24 Press, MC., Shah, N. and Stewart, G.R. I 19861in Biology and Control of Orobanche fter Borg, S.I., ed.), pp. 96-106, LHNPG Wageningen 25 Shah, N., Tuohy, I., King, G. and Stewart,

G.R. (1984) in Proceedings of the Third International Symposium on Parasitic Weeds (Parker, C., Musselman, L.I., Polhill, R.M. and Wilson, A.K., eds), pp. 74-80, ICARDA 26 Hunter, 1.1.and Visser, j.H. (1985) S. Afr. /. Bot. 52,81-84 27 Raven, LA. ( 1983) Adv. Ecol. Res. 13. 135-234 28 Wolswinkel, P. (1985) Physiol. P/ant. 65, 331-339 29 Kuijt, I. (1977) Annu. Rev. Phytopathol. 17, 91-l I8 30 Press, M.C., Tuohy, 1.M. and Stewart, G.R. ( 1987) in Parasitic Flowering P/ants (Weber, H.C. and Forstreuter, W., eds), pp. 631-636, University of Marburg Press 31 Stewart, G.R. ( 1987) in Parasitic Weeds in Agriculture I. Striga (Musselman, LJ., ed.), pp. 77-88, CRC Press 32 Shah, N., Smimoff, N. and Stewart, G.R. ( 1987) Physiol. Plant. 69,699-703 33 Ehleringer, j.R., Schulze, E-D., Ziegler, H., Lange, O.L., Farquhar, G.D. and Cowan, I.R. (19851 Science227,1479-1481 34 Govier, R.N., Nelson, M.D. and Pate, J.S. ( 1967) New Phytol. 66,285-297 35 Press, M.C., Shah, N., Tuohy, 1.M. and Stewart, G.R. (1987) P/ant Physiol. 85, 1143-1145 36 Mallabum, P.S., Press, M.C. and Stewart, G.R.I. Exp. Bot. (in press) 37 Schulze, E-D., Turner, NC. and Glatzel, G. ( 1984) Plant Cell Environ. 7.293-299 38 Press, M.C., Nour, I.]., Bebawi, F.F. and Stewart, G.R. (1989) /. Exp. Bot. 40 (in press) 39 de Hullu, E. ( 1984) in Proceedings of the Third In temational Symposium on Parasitic Weeds (Parker, C., Musselman. LJ.. Polhill, R.M. and Wilson, A.K., eds), pp. 43-52, ICARDA 40 Graves, I.D., Press, MC. and Stewart, G.R. (1989) P/antCe//Environ. 12,lOl-107

References 1 Kuijt, J.(1969) The BiologyofParasitic Flowering P/ants, University of California Press 2 Atsatt, P.R. (1983) in Encyclopedia of Plant Physiology (Vol. I2C New Series) (Lange, O.L., Nobel, P.S., Osmond, C.B. and Ziegler, H., eds), pp. 519-535, Springer-Verlag 3 Doggett, H. (1982) in Sorghum in the Eighties (Vol I ) (ICRISAT, ed.), pp. 3 13-320, ICRISAT 4 Last, F.T. ( 1960) Ann. Appl. Biol. 45, 207-229 5 Tsivion, Y. (1978) /srae/j. Bat. 27, 103-121 6 Klaren, C.H. and van de Dijk, S.I. (19761

263