Influence of halophytic hosts on their parasites—the case of Plicosepalus acaciae

Parasitic plants obtain nutrients, carbohydrates, ions and water from their host plants and develop special morphological features to survive on the living tissue of other plants. We investigated the mistletoe Psilocephalus acaciae on halophytic and non-halophytic hosts in the Arava Valley. When P. acaciae was parasitic on halophytic hosts, ions were transported by the transpiration stream in the parasite. Chloride and sodium induced the development of leaf succulence in the parasite. Increasing succulence is a morphological adaptation by the mistletoe to the increasing salt stress on halophytic hosts.


Introduction
Halophytes develop various physiological and morphological adaptations to cope with high salt concentrations in soil and water (Waisel 1972;Breckle 2002;Flowers and Colmer 2008), and the regulation of uptake of Na + , K + and Cl 2 and transportation and storage of these ions in plant organs are key issues to understand life strategies (Breckle 1990). In combination with the physiological processes, morphological adaptations like leaf bladders, salt glands and succulence develop to mitigate the effects of high salt concentrations in the photosynthesizing tissue by desalinization of leaves or dilution or compartmentalization within leaf tissues (Breckle 1974(Breckle , 2002Schirmer and Breckle 1982;Freitas and Breckle 1992;Veste 2007a).
Furthermore, the ion pattern and chemical composition of terrestrial halophytes are often characteristic of the species and family and only partly influenced by soil conditions ('physiotype concept'; Albert 1982). For example, some desert halophytic members of the Aizoaceae are characterized by a selective hyper-accumulation of Na + and Cl 2 in their leaves even on soils with low salinity Veste 2007a) as is also the case with the members of Amaranthaceae (Chenopodiaceae: see Veste and Breckle 2000;Albert 2005). On the other hand, members of the Poaceae are able to limit the uptake of Na + (Albert 1982(Albert , 2005. This emphasizes the importance of active ion regulation and selective transportation of ions already in the root system. An interesting question is what happens if the root system for such ion selection is missing, as in the case with parasitic plants? Is the pattern of ion accumulation between host and parasite mirrored to some extent? In this context, xylem-tapping mistletoes are a perfect model lacking a root system for selective ion uptake. Water and nutrient uptake is directly linked to their hosts (Glatzel and Geils 2009) through a haustorium, which connects the parasite with its host and allows the transportation of water, inorganic and organic compounds from the host's transpiration stream directly into the parasite. While there has been considerable attention to nutrient exchange between the parasite and the host (e.g. Lamont 1983;Glatzel and Balasubramaniam 1987;Popp 1987), only little attention has been given to mistletoes growing on halophytic hosts and the influence of the salinity experienced by the host on the parasite. There are several reports from mistletoes on mangroves: Phthirusa maritima (Loranthaceae) on Conocarpus erectus and Coccoloba uvifera (Goldstein et al. 1989); Lysiana subfalcata on Ceriops tagal (Ullmann et al. 1985), Loranthus rhamnifolius on Sonneratia alba and Loranthus sansibarensis and L. dregei on Lumnitzera racemosa (Walter and Steiner 1936) and Loranthus capitellatus (Holtermann 1907). In these cases, the mistletoes on mangroves have a 0.2-to 3.5-fold higher Na + concentration than their hosts (Lamont 1983;Goldstein et al. 1989). In South Africa and Namibia, of 19 mistletoes only Tapinanthus oleifolius (Loranthaceae) occurs on the halophytes Tamarix usneoides and Salvadora persica (Visser 1981;Popp et al. 1995), where it has a higher salt concentration in its xylem sap than the host. Tapinanthus oleifolius also increases its leaf succulence on the halophytic T. usneoides as well as on non-halophytic Euphorbia virosa (Veste 2007b), which is strongly correlated with the accumulation of inorganic ions (Popp et al. 1995). Tamarix ramosissima is an uncommon host for Phoradendron californicum (Loranthaceae) in North America, but the mistletoe is able to grow on the halophyte (Haigh 1996).
Another example for such a host -mistletoe association is Plicosepalus acaciae (syn: Loranthus acaciae; Loranthaceae) with an east-Sudanian distribution. The mistletoe occurs along the Arava Valley and some adjacent wadi systems north to the Jordan Valley and in some parts of Jordan (Shmida and Darom 1992;Qasem 2009Qasem , 2011. Within the Arava Valley, the highest density of P. acaciae can be found close to Yotvata (Todt et al. 2000), where most of the trees are parasitized by P. acaciae. The number of parasites decreases strongly a few kilometres north and south of Yotvata, and the movement pattern of the bird bulbul (Pycnonotus xanthopygos), as the main disperser of the red berries with sticky seeds, can explain the distribution of P. acaciae in the Arava Valley (Green et al. 2009). The most common hosts are Acacia trees because of their wide distribution (Munzbergova and Ward 2002;Bowie and Ward 2004). Other hosts for P. acaciae in the southern Arava Valley are trees and shrubs (Atriplex, Nitraria and Tamarix; Todt et al. 2000). On these hosts the parasite shows morphological and physiological plasticity as known for other halophytes (Breckle 2002) growing under different salt conditions. This high flexibility makes the different host -parasite associations of P. acaciae an interesting model system to understand adaptive plasticity to salt stress. We wanted to check how the parasite is able to cope with halophytic hosts growing under high salinity stress and the ecophysiological plasticity of ion accumulation that is involved. A special focus is given to ion uptake and its link with the development of leaf succulence at different growth stages on different hosts. We also compiled an updated checklist of the known hostparasite associations.
Statistical analyses were performed with SPSS package 7.5.1. We checked the normality of distributions and variance of all samples with the Kolgomorov -Smirnow test (Kö hler et al. 1996;Lozá n and Kausch 1998). For correlation analysis we used Spearman's rank correlation coefficient (r's). Significance of differences between samples fixed at the 0.05 probability level for all statistical tests was checked with the Mann -Whitney U test.

Results
A list of the hosts of P. acaciae at the study site, Yotvata, in the Arava Valley and some other areas of the Middle East are given in Table 1. A high proportion of wild host species had reduced leaves and photosynthesis occurred mainly in stems. The hosts Haloxylon persicum and Calligonum comosum grew on sand dunes in Yotvata and other sandy areas in the Arava Valley. Albizzia, Casuarina and Delonix are introduced species and grew within the Yotvata settlement, whereas Tamarix aphylla is native to the northern Negev, but is planted in Yotvata as shelterbelts around fields and date palm plantations. The size of the mistletoe varied between the different hosts. The largest (and most) individuals of P. acaciae grew on Acacia and Tamarix, while only a few (and smaller) individuals occurred on Atriplex and Haloxylon. The highest number of P. acaciae in the area was on Acacia tortilis, A. raddiana and Tamarix nilotica, with no observable differences in density or preferences for non-halophytic or halophytic hosts.
We determined leaf area (Fig. 2), leaf thickness (Fig. 3) and thus leaf volume (Fig. 4) for P. acaciae, which varied according to host and increased considerably with leaf age. In old leaves, leaf volumes were up to 3-fold higher when parasitizing halophytic hosts in comparison with non-halophytic hosts. Leaf area, thickness and volume increased very significantly for P. acaciae on all Tamarix species (Figs 2-4) when compared with non-halophytic hosts. The maximum leaf area of P. acaciae on T. nilotica was 17.0 cm 2 (median 13.0 cm 2 ), whereas on A. tortilis the maximum leaf area was 8.1 cm 2 (median 6.01 cm 2 ). The leaf thickness of old leaves ranged from 1. thickness and succulence was also observed for old leaves of P. acaciae growing on the halophyte Nitraria retusa compared with the non-halophytic species, except that the differences were not significant on Calligonum cumosum and Casuarina cunninghamiana. Succulence, expressed as water content related to organic dry matter (Fig. 5) The ash content, which represents the sum of ions accumulated, exhibited much higher values in the parasites growing on halophytic than non-halophytic hosts (Fig. 6). The ash content of old leaves of plants growing on T. nilotica was 30.6 % compared with 12.1 % on Acacia tortlis. Accumulation of Na + was similar to that of Cl 2 at higher concentrations (Fig. 7). At lower concentrations, the uptake of Na + could be very low and seemed to be better controlled than Cl 2 uptake (Fig. 7). The uptake of Na + and K + was not strongly correlated but still definitely antagonistic (Fig. 8). Therefore, it follows that K + uptake was also antagonistic to Cl 2 (Fig. 9). Succulence of leaves of  P. acaciae increased significantly with increasing leaf Na + concentration (Fig. 10), and with an even stronger correlation if both Na + and Cl 2 were taken into account (Fig. 11). Our next question was whether the ionic pattern of the parasite reflected that of the host? This we checked by comparing the ion ratios in the parasite and the hosts. The increase of Na + on halophytic hosts could also be deduced from the higher Na/K ratio on halophytic than nonhalophytic hosts (see Table 2). Our samples also showed that the soluble fraction of Ca 2+ (the hot water fraction, which is approximately the portion that is osmotically relevant) was very different between species (Table 3), where the two chenopod hosts caused much higher Ca 2+ concentration in the parasite than in the other species (Table 3). The results indicate that the host's concentrations of osmotically active ions are not strongly correlated with the equivalent ions in the parasite (Table 4 and Fig. 12).

Discussion
Previously published information on hosts of P. acaciae (Zohary 1966;Feinbrun-Dothan et al. 1991;Shmida and Darom 1992) mentioned Atriplex and Tamarix but no   Table 1). In total, there are now 37 non-and 6 halophytic hosts known. Similar to other halophytes, the succulence observed for P. acaciae growing on halophytic hosts showed a morphological adaptation of the mistletoe to salt stress (Breckle 1990). The increased succulence presumably resulted from absorption of Na + and Cl 2 ions from the host and osmotic adjustment by the parasite (Veste and Breckle 1995). There was a highly significant correlation between the succulence and Na + and Cl 2 accumulation, which is in accordance with observations made by Popp et al. (1995) for T. oleifolius growing on Tamarix usnoides in Namibia. An increase of Na + can also be deduced from the higher Na/K ratio on halophytic     Table 2), a very general rule for many other non-parasitic halophytes (Albert 1982;Reimann and Breckle 1993;Breckle 2002). Popp and Richter (1998) also reported an increase in Na/K ratios for T. oleifolius growing on T. usneoides from 1.58 to 13.21 with increasing leaf age.
The concentrations of osmotically active ions were not strongly correlated with those of the parasite. This may easily be explained by the fact that the samples of leaves from the parasite and of the corresponding young host stems or host leaves were not directly osmotically dependent. Similarly, the Na/K ratios did not reflect the relation between the host and the parasite (Tables 2 and 4).
It is difficult to be certain about the concentrations of ions available to the haustoria of the parasite, as obtaining xylem sap (the presumed source) is in itself difficult. However, the concentrations of monovalent ions in the leaves of the host are unlikely to provide a good indication of the xylem concentrations due to the presence of salt bladders (Atriplex species) or salt glands (Tamarix species). Additionally osmotic adjustment may involve compatible solutes (Yancey 1994;Flowers and Colmer 2008), which we did not investigate; they are, however, found in other mistletoes (Popp and Polania 1989), although in the more succulent leaves of T. oleifolius they are of minor importance for osmotic adjustment (Popp et al. 1995).

Conclusions
Various definitions and classifications of halophytes have been proposed over the past decades, but following the   classification of halophytes by Breckle (1990Breckle ( , 2002, which includes the mechanisms for controlling the NaCl concentration in plants, P. acaciae is a facultative eu-halophyte. Plicosepalus acaciae increases its halosucculence according to the host, which is in this respect comparable with soil substrates. In any case, the water and nutrient status of the host plants, especially under desert conditions has a strong influence on the equilibrium of host and parasite and thus may mutually influence mortality (Bowie and Ward 2004). Surface structure, size and indirectly host origin influence the success of germination and of seedling establishment of parasite on host twigs (Rödl and Ward 2002): whether, during these early stages of development, the halophytic character of the host plant and salinity plays a role is still unclear. Host specificity seems to be rather low in mistletoes (Norton and Carpenter 1998;Norton and de Lange 1999)