Chlorophyll content and fluorescence as physiological parameters for monitoring Orobanche foetida Poir. infection in faba bean

Orobanche spp. are root parasitic plants that cause yield losses in faba bean (Vicia faba L.). In Tunisia, O. crenata and O. foetida are among the major problems limiting faba bean production and productivity. Breeding for resistance and development of resistant varieties remain the most efficient control strategy to combat these parasites. In our study, field trials were conducted over two consecutive cropping seasons. A set of 42 genotypes were used in this study; 39 advanced lines and three checks; Najeh and Baraca (resistant) and Badi (susceptible). The trials were conducted in highly infested O. foetida plot at Oued-Beja Research Station in Tunisia. Results showed that advanced lines XAR-VF00.13-1-2-1-2-1 and XBJ90.04-2-3-1-1-1-2A expressed high resistance level exceeding those recorded for resistance checks Najeh and Baraca. O. foetida significantly affected the biomass, grain yield, chlorophyll content index (CCI) and the maximum quantum efficiency (Fv/Fm ratio). No significant effect was observed on host plant water content (WC). CCI decreases varied from 46.4% for the susceptible check Badi and 4.2% and 9.3%, respectively, for Baraca and XBJ90.04-2-3-1-1-1-2A. Orobanche parasitism resulted in a slight decreases of Fv/Fm ratio for the advanced lines XBJ90.04-2-3-1-1-1-2A and XAR-VF00.13-1-2-1-2-1 against important decreases observed for Badi and Baraca. Correlation between resistance to O. foetida and CCI and Fv/Fm makes this, easy-to-measure, parameter very useful as a practical screening tool for early parasitism detection, diagnosis and identification and selection of high resistant plants against this parasite.

considered as a serious threat, they cause important damages and yield losses on many legume crops [1][2][3]. In Tunisia, O. foetida, O. crenata, O. cumana, and Phelipanche ramosa are parasitizing many crops such as faba bean, chickpea, lentil, grass pea and sunflower [4,5]. While O. crenata was mentioned as a serious pest for decades, O. foetida has been presented as an emerging problem for many legume crops such as faba bean, chickpea, lentil, grass pea, medick, common and narbon vetch [4,6,7]. The Orobanche infested area in Tunisia is estimated to more than 80,000 ha mostly situated in the main grain legumes production area (unpublished data, authors' own estimates). In high infested fields, farmers abandoned planting legumes especially faba bean and switched to non-host crops such as wheat leading to a strict wheat mono-cropping system. The devastating effect of Orobanche is associated with their multiplication rate and high seed viability (15-20 years) [8]. Many control methods were used to control Orobanche, including agricultural practices, chemical and biological methods [9,11,14]. The chemical method using especially synthetic strigolactones and herbicides were the most explored but resulted in a limited success due to their application complexity in the farmer's field [8]. Till date, no single control method has shown successful with full control of the parasite. All control strategies resulted in an incomplete protection of the crop [9][10][11]. To date, the only effective method is through an integrated management strategy with genetic resistance as a main component. Farmers should use resistant varieties, avoid planting contaminated seeds and follow preventive practices to limit the spread of the parasite to new fields [12]. While avoidance of broomrape dispersal is virtually difficult, crop resistance and prevention measures could be the most effective and economical methods to reduce this root parasitic weed infestations. Genetic resistance coupled with other control methods resulted someway in good control of the parasite with significant decreases of the damages. At this level, research is needed for generating new technologies and developing new resistant varieties and effective screening tools. Many resistance mechanisms were studied focusing mainly on the physical and biochemical host-parasite interface such as low production of Orobanche seed germination stimulants and/or release of inhibitors by the host root system [2,13,14], the existence of a host plant roots physical barrier resulting in unsuccessful haustorial penetration and necrosis [15] and the development of a deep root system that escape Orobanche infestation [16,17]. In addition, an integrated control strategy could be improved through early detection and monitoring of the underground infestation and the parasite development. Rousseau et al. [18], mentioned that plant infestation by root parasitic weeds has a systemic impact that could be observed on host leaves. In this regard, chlorophyll fluorescence, which is a non-destructive and rapid assessing tool of photochemical quantum yield and photoinhibition, could be used for early Orobanche infestation and estimate its impact on the host plant. It is widely used as a plant response indicator under biotic and abiotic constraints such as heat, drought, waterlogging, salt stress, nitrogen deficiency, pathogen infection and herbicide resistance [19][20][21].
However, only few studies were conducted on parasitism effect on host plant chlorophyll fluorescence [22][23][24]. As reported by Maxwell and Johnson [25], the photochemical processes alterations are usually the first signs in the stressed plant leaves that could be used to estimate photosynthetic performance under stress conditions. These photochemical processes alterations appear in the chlorophyll fluorescence kinetics and induce changes in the established fluorescence parameters and consequently PSII damages. Cameron et al. [26], reported that the parasitic plant Rhinanthus minor significantly reduced biomass production in Phleum bertolinii and demonstrated that such decrease was reflected by changes in photosynthetic activities and significant reductions in the quantum efficiency of PSII and chlorophyll concentration.
In this paper we aim to evaluate the performances of faba bean advanced lines collection under O. foetida infested conditions and assess the impact of the parasite on plant growth and seed yield in correlation with physiological behavior using chlorophyll content and chlorophyll fluorescence parameters.

Plant material and field trials
Genotypes evaluation and screening for resistance to O. foetida. A set of 39 smallseeded faba bean advanced lines, developed from crosses performed in Tunisia (Table 1), were used for a first-year (2009/2010) screening and evaluation for resistance to O. foetida. Three checks were added to the list, two Tunisian varieties Badi and Najeh and a Spanish variety Baraca. Both varieties Najeh and Baraca, carrying partial resistance to O. foetida and O. crenata [1,27] were used as resistance check while Badi was used as susceptible check. The screening was performed under high O. foetida infested sick plot at Oued-Beja Research Station-Tunisia (36˚44'N; 9˚13'E). The trial was conducted in a randomized complete block design with three replications. Each genotype was planted, at a density of 24 seeds per m 2 , in four rows of 4 m length and 50 cm inter-rows spacing. The planting was performed the last week of November. No fertilizer's supply or herbicide treatments were applied after plant emergence, only hand weeding was carried out keeping only faba bean plants and emerged Orobanche shoots.
Confirmation of the resistance and assessment of the parasitism impact. Out of the total tested collection, the two best resistant genotypes XAR-VF00.13-1-2-1-2-1 and XBJ90.04-2-3-1-1-1-2A were selected all with the three checks to conduct the second-year (2010/2011) evaluation and confirmation trials. These genotypes were selected not only based on the cropping season (2009/2010) but also from preliminary data and observations recorded during previous cropping seasons (data not shown). The five genotypes were planted the last week of November in both infested and free Orobanche fields. Both trials were conducted same as described in the first-year screening.
For both cropping seasons, monthly rainfall and average temperature distribution for the two cropping seasons collected from the iMETOS meteorological station (Pessl instruments) are presented in the Table 2.

Measurements
The field evaluation of the studied genotypes and assessment of their resistance level to O. foetida was evaluated through different parameters measured at different crop development stages.
During the first-year screening, data related to Parasitism Index (PI), number of emerged Orobanche shoots (EOS) per faba bean plant and seed yield (SY) (g.m -2 ) were recorded at harvesting time on the two central rows.

PI ¼ ðOIN�OSVÞ=100
• OIN: Orobanche incidence or percentage of plants showing at least on Orobanche emerged shoot • OSV: Orobanche severity (1-9 scale) or level of damage caused by Orobanche on the host plant development and seed production [28].
During the second-year evaluation, in addition to OIN, OSV, PI, EOS and SY mentioned above, the number of days to Orobanche emergence (NDOE) and total emerged Orobanche number per plant were also recorded. At pod-setting stage and from both infested and noninfested fields, five random faba bean plants from each plot were carefully dug-up with all Table 1. Origin and main characteristics of different studied genotypes.

Genotypes
Cross/Pedigree and main characteristics Orobanche attachments. For each single plant the biomass and water content (WC) were recorded and Orobanche attachments (TON) were classified into emerged and non-emerged attachments.
Chlorophyll content index (CCI) and the maximum quantum efficiency (F v /F m ratio) measurements were recorded once a week between 10 am and 1 pm from 123 to 144 days after planting (DAP) in both infested and non-infested fields. The two parameters were recorded on leaves from the 11 th main stem node of random faba bean plants. CCI was measured on five random faba bean plants per plot using an CL-01 Chlorophyll Content Meter (Hansatech Instruments Ltd, UK). For every measurement almost the same part of the leaf was placed between two clips and the CCI was determined in dual wavelength optical absorbance (620 and 940 nm).

Statistical analysis
The statistical analysis, ANOVA and means comparison, were performed using the SPSS statistical program v.21. Differences among treatments for all measurements were compared at P = 0.05 and using Duncan's multiple-range test.

Field evaluation and identification of potential resistance genotypes to O. foetida
Results showed high variability in the resistance to O. foetida between the genotypes. Significant differences were observed for the EOS, PI and SY (Fig 1). Based on PI, almost 46% of the tested genotypes showed a resistance level higher than the resistant check Najeh. The two advanced lines XBJ90.04-2-3-1-1-1-2A and XAR-VF00.13-1-2-1-2-1 expressed a high resistance level to O. foetida with respective PI of 1.2 and 2.2. Both genotypes showed low Orobanche infestation level with only 0.9 and 1.4 EOS, respectively. Such resistance observed for these two genotypes was reflected by a high seed yield with 154.2 g.m -2 and 257 g.m -2 respectively. They produced almost two (1.8) and three (2.9) times more than the resistant check

Number of days to
Chlorophyll content index and chlorophyll fluorescence. Orobanche parasitism significantly affected (P � 0.01) the host plant chlorophyll content index (CCI) and F v /F m ratio for all the studied genotypes (Figs 4 and 5). All genotypes showed significant difference between infected and non-infected plants before Orobanche emergence. Between 125 and 146 DAP, CCI decreases under O. foetida infestation varied from 23.6% for Baraca to 77.2% recorded for Badi. Respective decreases of 19.4% and 30.8% were recorded for the same genotypes under free Orobanche field. Clear differences were observed for the CCI between infected and free Orobanche plants. Such differences were more pronounced for the susceptible check Badi. Variation between 125 and 146 DAP, varied from 4.2% for Baraca to 46.4% observed for Badi.
F v /F m ratio was significantly affected by Orobanche parasitism (Fig 5). In infected plants F v / F m decreased by 58.8% (0.789 to 0.325) for the susceptible check Badi against only 9.9% (0.787 to 0.709) observed for free Orobanche plants. Decreases of 46.2%, 14.5%, 5.9% and 4.7% were Table 4

Discussion
Results from the first-year screening showed high variability for the resistance to O. foetida in the tested collection. Resistance to broomrapes is not only the capacity of the genotype to limit the parasite development and the damages that causes, but also the capacity of that same genotype to grow and produce grains under such parasitism attack. Two advanced lines, XBJ90.04-2-3-1-1-1-2A and XAR-VF00.13-1-2-1-2-1 were identified and selected for their high resistance level and good SY under high O. foetida infestion. Both genotypes showed a low

PLOS ONE
Faba bean (Vicia faba L.) response to Orobanche parasitism infestation level compared to resistant and susceptible checks. Previous studies showed that Orobanche causes early wilting symptoms in parasitized plants and results in a shortening of the reproductive phase and affecting significantly the flowering, pod setting, plant biomass and seed production [13,16,28]. Such effects were highly pronounced for cv. Badi, for which O. foetida has severely restrained plant growth, and resulted in almost complete damage and yield losses. Moderate effect of the parasite on plant development and seed production was observed for both selected advanced lines XAR-VF00.13-1-2-1-2-1 and XBJ90.04-2-3-1-1-1-2A. Results also showed that despite the biomass decreases recorded for different studied genotypes, no significant effect of Orobanche parasitism was observed on the host plant's WC. Indeed, due to the parasitic burden and resources sinking the host plant has limited its biomass and dry matter production and allocation in order to keep its physiological functioning through a normal and optimum water content. It was also reported that Orobanche parasitism effects on host plant growth and biomass production and allocation are directly associated with the infestation level [2,9,10]. Ennami et al., [29] reported a high negative correlation between both faba bean and lentil plant's growth and O. crenata development attributing that to the competition between the host and the parasite for nutrients. Other previous studies reported that the detrimental effect of both O. foetida and O. crenata on faba bean grain yield

PLOS ONE
Faba bean (Vicia faba L.) response to Orobanche parasitism can reach up to 90-100% depending on the infestation severity and the broomrape-crop association [30,31]. In addition, our results showed significant differences in CCI between infected and noninfected plants for the five tested genotypes, even before Orobanche emergence. These results indicate that this parameter could be very useful for early detection of the underground Orobanche infestation. In fact, several studies reported the importance of the number of host leaves and their greenness in plant eco-physiological studies because they provide information about physiological responses of plants under stress conditions [32,33]. Decreases in CCI that were observed in parasitized plants could be explained by the parasite nutritional requirements that limits the normal growth and functioning of the host plant. Similar results were reported for tomato/P. ramosa pathotype [34,35]. Shamsullah et al. [35], found that, compared to noninfected plants, P. ramosa decreased chlorophyll content in tomato leaves by 29.17%. Shen et al., [23] reported also similar results with Mikania micrantha/Cuscuta campestris and showed, also, that despite the CCI decrease observed on the M. micrantha leaves, there was no significant effects of C. campestris parasitism on chlorophyll a:b ratio.  [34] who showed that F v /F m , which is proportional to the PS II quantum yield and well correlated with the photosynthesis quantum yield [36], was significantly decreased by P. ramosa attack on tomato plants. In the same study, the authors demonstrated that the F v /F m reduction is mainly induced by an effect on the variable fluorescence (F v ) resulting in PS II electron transport damage. Similarly, Rousseau et al. [18], reported that F v /F m in Arabidopsis thaliana leaves was impacted by P. ramosa. Also, other studies showed that C. reflexa induced a sink-dependent stimulation of net photosynthesis on Ricinus communis [37] and that such infestation by C. campestris decreases host stomatal conductance, transpiration, chlorophyll content, and soluble protein concentration on M. micrantha [23]. These results are contrasting with other studies [38,39] who reported that broomrape affects host biomass and yield and related traits with only minor disturbance in host's leaves tissue but no perceptible effects on photosynthetic rate. More recently, Ennami et al. [29] showed that effective quantum yield of open photosystem II, (Fm'-F)/Fm', was significantly reduced by O. crenata attack on susceptible faba bean and lentil genotypes. The effects caused cause by the parasitic weeds on the different parameters may directly or/and indirectly affect the functioning of the photosynthetic system and rate and therefore affect the growth of the host plant.

Conclusions
O. foetida can affect faba bean host plants in/through different ways and at a big range of scales, from the root to the leaves through the whole plant. Out of the initially larger tested faba bean collection, the two genotypes XAR-VF00.13-1-2-1-2-1 and XBJ90.04-2-3-1-1-1-2A expressed the highest resistance level to O. foetida and showed a moderate and limited effects of the parasite on plant development and seed production. For both genotypes, the high resistant level was reflected by limited effects of the parasite on plant growth, biomass production, seed yield and physiological functioning of the host plants. Significant variations in CCI and F v /F m were observed from individual plants between the tested genotypes and between infected and noninfected plants. The significant positive correlation observed between CCI, F v /F m and high resistance level to Orobanche may suggest the integration of these physiological traits in plant selection and screening for resistance to broomrapes. These practical screening tools could be coupled with other new smart imaging technologies for early detection of the root parasitic weeds infestation.