Seed Germination Characteristics of Rhus tripartitum (Ucria) Grande and Ziziphus lotus (L.): Effects of Water Stress

Ziziphus lotus (L.) Lam. (Rhamnaceae) and Rhus tripartitum or Sumac (Anacardiaceae) are two indigenous species from arid and semiarid regions of Tunisia, characterized by a severe climate where dry seasons are very long. The combined action of anthropogenic factors and climate in arid regions caused a gradual threat of plant assets. In this context, an experimental study of the effects of water stress (0 to −1MPa) on seeds has identified the water requirements germinal stage of both species. The results showed that both species were able to germinate at relatively low water potentials. However, beyond −0.6MPa, germination was completely inhibited for R. tripartitum, when it reached for another 50% for Z. lotus. Increasing the concentration of PEG 6000 progressively inhibited germination in both species. Only Z. lotus could be considered tolerant of water stress, because, to −1MPa, seeds germinated with a rate of 17%. It resulted in that the species Z. lotus presented an adaptive capacity to aridity much greater than that observed for R. tripartitum.


Introduction
Drought stress is considered to be the main environmental factor limiting plant growth and yield of many agronomic and horticultural crops, especially in semi-arid areas. In Mediterranean-type ecosystems, seasonal water shortage is the main factor constraining survival and growth of plants.
Soil depth and texture are considered the most important edaphic properties that influence the moisture regime in arid environments with episodic rainfall. To date, a great deal of effort has been focused on physiological process underlying plant responses to drought stress.
The current situation of arid and desert areas of Tunisia (three quarters of the area) is in rapid decline of natural vegetation cover associated with an erosion of biodiversity [1,2]. This decline is attributed to particular stressful environmental conditions, land clearing, and overgrazing, resulting in effects of increasingly adverse ecological (desertification) and economical.
In this area, the steppic vegetation is dominated by tall perennial grasses [3]. But, sparse trees and shrubs are among the most threatened species because of their excessive use for domestic purposes and their poor regeneration performances [4,5]. On the other hand, under arid bioclimate, the flora has been subjected to a high, permanent increase of human pressure since at least the last century. Such a situation contributed and might have induced the phenomenon of desertification [5,6].
In Tunisia, several attempts have been made to restore degraded rangelands [7][8][9]. The improvement of the methods based on seed dispersal and seed sowing would enable to know more about seed response to main environment factors especially drought stress in arid land. This factor could influence germination parameters [10,11]. Under these conditions, seed germination monitoring in relation to water is very important to determine the colonisation capacity of species [12].
In this area, several native species are potentially interesting under aspects of dune stabilization and extension of plant cover, including some Ziziphus species and Rhus tripartitum. In addition, they improve the stability of ecosystems where  they are present, contribute to reduce the risks of desertification, and are helpful in restoring degraded ecosystems [6,[13][14][15][16]. These shrubs are also a feeding resource for livestock especially during the summer, when the alternative herbaceous species have wilted [17]. According to Barroso et al. [18], these species relieve sand movement. Currently, Z. lotus L. and R. tripartitum are rarefied in this area under overgrazing and their artificial reintroduction is necessary for the restoration of degraded ecosystems.
Accurate knowledge on the germination requirements of Z. lotus L. and R. tripartitum species is now required for successful uses of these two species in operations of artificial regeneration in Tunisia. Therefore, the effects of low water availability, simulated by PEG 6000 , on seed germination were studied for these two tree species located in arid and semiarid areas of Tunisia. temperature ranges between 3.9 ∘ C throughout the coldest month (January) and 36.2 ∘ C throughout the hottest month (August).

Germination Experiments
where is the concentration of PEG 6000 in g⋅L −1 of water, is the temperature in ∘ C, and Ψℎ is osmotic potential in bars. Distilled water served as a control. Seeds of the same form for both species (Z. lotus and R. tripartitum) have been selected and sterilized by aqueous solution of 75% ethanol for 5 min to prevent fungal attack and rinsed in several changes of sterile distilled water. The seeds were germinated in 9 cm sterile Petri dishes lined with two sterile Whatman number 1 filter papers with 5 mL of distilled water or the respective test solutions [20]; there were 20 seeds per Petri dish and five replicates in each treatment. Germination tests were conducted in complete darkness at 30 ∘ C. Distilled water was added into the Petri dishes each day to maintain the concentration of the test solutions. Seeds were considered to be germinated when the radicle emergenced. After seed sowing, the germinated seeds were counted and eliminated every second day for 21 days.

Methods of Germination Expression.
The characteristics of seedling emergence were (FG), number determined: final germination percentage of days to first germination (delay of germination), critical limits of germination and mean time to germinate (MTG).
(i) Final germination percentage (FG) was calculated as the cumulative number of germinated seeds with normal radicles [19]. GP = ∑ , where is the number of seeds that had germinated at each counting. (ii) Mean Time of germination: MTG was estimated according to the formula: where is the number of seeds germinated at day , the incubation period in days, and the total number of germinated seeds in the treatment [21].

Statistical Analysis.
Statistical analysis was performed using two-way analysis of variance (ANOVA and DUNCAN) to test the effects of water stress, species, and their interactions, on germination characteristics. All statistical methods were performed using SPSS, version 16.

Influence of Water Stress on the Kinetics of Germination.
The kinetics of germination by osmotic stress conditions is presented in Figure 2. It reflects the sensitivity of the species to water stress. In control condition, the kinetic curves show three phases of germination: latency, accelerating exponentially, and finally step corresponding to a stop germination after attaining the maximum germination. The depressive effect of water stress on germination occurs during one or all of these three phases, depending on the degree of the lowering of water potential and species studied.
It results in a slower processing speed visible from −0.2 MPa, and thereafter increases. Whose, the delay of germination the most extensive (3 days) is obtained to −1 MPa for Z. lotus while 7 days to −0.6 and −0.8 MPa for R. tripartitum.

Final Germination Percentage.
Final germination percentage of Z. lotus seeds were recorded in controls with (100%) and only 66% of R. tripartitum. The germination capacity is greatly affected by water stress (Figure 3), analysis of variance showed that there is a highly significant difference between treatments and species ( < 0.0001) ( Table 1). Test ANOVA with two factors, calculated at 5% level, presented a treatment effect and species very highly significant ( < 0.0001) and the interaction between them (Table 1).
Duncan test (Table 2) discriminates two homogeneous groups under species effect: G1 = A to Z. lotus and G2 = B to R. tripartitum.
Duncan's expresses homogeneous groups according to the order of decreasing water potential (Table 3). It is found that upon application of −0.2 MPa, the germination capacity is significantly reduced especially for R. tripartitum where it exceeds 50%.

Critical Limits of Germination.
Critical limits of germination of both species were highly different. Seed germination of Z. lotus vanishes up there from −1 MPa but at −0.8 MPa for R. tripartitum. So, the critical limit of germination of Z. lotus are the order of −1 MPa while it was −0.6 MPa for R. tripartitum (Figure 3).

Influence of Water Stress on the Final Percentage and Mean Germination Time (MTG).
To better understand the effect of water stress on germination of Z. lotus and R. tripartitum, we showed the relationship between mean germination time and the final percentage. In fact, in Z. lotus, a water potential of −0.2 MPa causes a decrease in germination capacity of around 30%. This decrease becomes important to −0.8 MPa (76%) and even more accentuated at −1 MPa (83%).
The results obtained for GMT of Z. lotus range from ≈5 to 7 days respectively, for 0 and −1 MPa, whereas Rhus tripartitum, varies from 9.67 to 11 days, respectively, to 0 and −0.6 MPa. So, the lowering of the water potential leads to an increase in the mean germination time.
The comparison of both species responding to water stress shows that Z. lotus is the most effective (tolerant) for most treatments compared to R. tripartitum (Figures 4 and 5).

Discussion
The main objective of our paper was to assess the germination characteristics of Z. lotus and R. tripartitum under abiotic constraints. The most important result showed that germination capacity of studied species under the environmental constraints and factors is sufficiently assured to consider these species for a reforestation program and the extend consequently their area of distribution. Additionally, studied species responded differently to stress induced by PEG 6000 . However, Z. lotus manifested better adaptation to drought conditions compared to R. tripartitum.
Under arid bioclimate, the successful establishment of plants largely depends on the success of germination. The  success of shrubs under warm and dry conditions is primarily dependent on optimal conditions (25/35 ∘ C and 12 h night/12 h day) for germination and recruitment [22].
Our results suggested that germination of Z. lotus seeds continued despite the increase of osmotic potential (Table 1). Up to −0.6 MPa, germination rate was superior to 50% for Z. lotus but completely inhibited for R. tripartitum ( Table 1). The higher germination percentage of Z. lotus compared to R. tripartitum may be related to its better adaptation to water deficits. Our results are consistent with the results of Ne' eman et al. [23] who showed that germination of Rhus coriaria was reduced by 80% at water potential of −0.26 MPa. In the same way, Noumi et al. [24] showed that up to −0.7 MPa seeds germination of R. tripartitum is totally inhibited.
Under controlled conditions, the highest germination percentage of studied species was obtained for non-stressed seeds. The variation of PEG 6000 concentration has significantly affected germination rate.
Overall, our results showed that the increase of water stress introduces the decrease in germination rate of studied species. This finding is consistent with the results of Ne' eman et al. [23] who showed that germination of Rhus coriaria was reduced by 80% at water potential of −0.26 MPa. In the same way, Chuang et al. [25] showed that the critical value of Periploca sepium was −1 MPa and the maximum value was −1.  [26].
In contrast to our results, the critical threshold of germination of Z. lotus and R. tripartitum is very highest compared to others species growing under arid and Saharan climates such as Lotus creticus (−0.7 MPa), Plantago albicans (−0.5 MPa), and Rantherium suaveolens (−0.3 MPa) [27]. In the same context, Sharma [28] showed that germination of Atriplex vesicaria Heward and Atriplex nummularia Lindl In Tunisia, several attempts have been made to restore degraded rangelands [7-9, 13, 14]. One important technique for the rehabilitation of degraded ecosystems is to improve the dispersal of seeds, which is labour saving when compared to transplanting nursery seedlings [15]. Arid zones in Tunisia, which cover more than 70% of the total area [2], are characterized by drought and high temperature and more or less general salinity. In addition many shrub species such as Z. lotus, R. tripartita, Periploca angustifolia, Retama raetam, Acacia raddiana, Ceratonia siliqua, and Lycium shawii are valuable species for afforestation in arid area of Tunisia as they withstand the harsh climatic conditions, improve the stability of ecosystems where they are present and contribute to reduce the risks of desertification, and are helpful in restoring degraded ecosystems [6,[13][14][15][16].
In the field, establishment of Z. lotus and R. tripartitum varies according to the season (cool/warm) and the geographical and ecological distribution as well as the amount of rainfall. Seasonal fluctuations of rain in deserts of Tunisia explain the variable establishment pattern of this plant during the course of a year, governed by seasonally different temperatures and the rainfall events. From the present results, it is evident that germination percentage decreases with increasing water stress and comes to the limit at −1 MPa.

Conclusion
Some demographic studies [29,30] suggest that the establishment of many woody species from arid and semi-arid areas is a rare event, which is mainly due to the occurrence of excessive water stress caused by low rainfall and/or high evapotranspiration rates. However, the establishment of these woody species would be possible only after years of above normal and well-distributed rainfall [31].
In conclusion, Z. lotus and R. tripartitum species manifested different levels of adaptation to drought. Increasing the concentration of PEG 6000 progressively inhibited germination for both species. Only Z. lotus can be considered tolerant to water stress, since, to −1 MPa, seeds germinate at a rate of 17%. As a result, Z. lotus presents an adaptive capacity to aridity much larger than that observed for R. tripartitum. Therefore, these findings form the basis for future trials involving the use of indigenous shrubs in the restoration of rangelands.