The Effects of Salt Stress on Germination, Seedling Growth and Biochemical Responses of Tunisian Squash (Cucurbita maxima Duchesne) Germplasm

Salt stress is considered as one of the most common abiotic stresses reducing the productivity and fruit quality of crop plants. The present study was carried out to assess the salt tolerance among 15 local squash (Cucurbita maxima Duchesne) landraces. Different salt (NaCl) concentrations of 0, 100, 200 and 300 mM were selected in order to evaluate the response of the study germplasm to salt stress based on 12 agronomic parameters and 3 biochemical traits, proline, malondialdehyde (MDA) and chlorophylls. A varied effect of the salt stress level was observed among the studied landraces based on germination potential, as well as on growth and biochemical parameters at seedling stage. Results showed that all landraces were drastically affected at high stress level with a significant variation in their stress response, indicating the existence of considerable genetic variability. Landraces “746” and “747” were the best performing cultivars across stress levels, whereas “1007”, “1008” and “1009” were the most negatively affected. Based on the tested landrace performance, four landraceswere selected and further evaluated at biochemical level, focusing on the determination of compounds that play a key role in the ability to withstand salt stress. The mean MDA content across landraces was generally increased in stressed plants, as compared to the control treatment; the increase was attributed to a peak in MDA content at specific stress levels. In particular, “746” and “1007” showed the maximum content at 100 mM NaCl, while in landrace “751”, MDA content reached its peak at 300 mM NaCl. In addition, the response of most landraces to salt stress involved an increase in free proline content, with the exception of “746”, with the maximum content being observed either at 200 mM (“748” and “751” landraces) or at 300 mM NaCl, where only “747” expressed the highest content. These findings can be extrapolated into efforts to develop more salt-tolerant squash landraces and exhaust the possibilities of using saline water or soils under changing climate conditions.


Introduction
The genus Cucurbita, belonging to the family Cucurbitaceae, includes several economically and nutritionally important vegetable crops cultivated worldwide [1]. It contains five domesticated species, namely C. argyrosperma Huber, C. ficifolia Bouché, C. maxima Duch., C. moschata Duch. and C. pepo L. [2]. Cucurbita maxima Duch. is an allogamous and extremely diverse species that originated in South America from the wild progenitor C. maxima ssp. andreana (Naud.). The species contains a series of squash ecotypes, whose deteriorating fruit quality. Despite the acquired knowledge in relation to the salt stress response in a plethora of plant species [33,38,39], there is a gap in relevant research fields for squash. Considering that the development of salt-tolerant germplasm is one of the most effective means for enhancing squash production in saline soils, this study aimed at investigating the response of squash germplasm to salt stress at the germination stage and examining the potential of selecting salt-tolerant genotypes at the laboratory level early. As such, 15 Tunisian squash landraces were subjected to NaCl-induced salinity stress at varying stress levels (0, 100, 200 and 300 mM of NaCl), and their response was assessed on the basis of traits related to seed germination and seedling growth potential. Moreover, based on the results related to salt stress response, four landraces were selected and further assessed in terms of biochemical parameters that are routinely employed as indicators of salt tolerance, namely malondialdehyde (MDA), free proline and chlorophyll a and chlorophyll b content.

Effect of Salt Stress Level on Traits Related to Seed Germination and Seedling Growth Potential
In order to evaluate the response of 15 Tunisian squash germplasm to salt stress (see the corresponding Table in Section 4), seeds were germinated on solutions differing in NaCl concentration (0, 100, 200 and 300 mM), and seed and seedling parameters were evaluated based on 12 criteria (see the corresponding Table in Section 4). The obtained data revealed a significant effect of the salt stress level (p < 0.001) on the germination and seedling growth potential of the tested squash landraces, with the stress effects being in most cases analogous to the stress level applied, therefore leading to most drastic effects at high stress levels for all the traits under study (Tables 1 and 2). S.O.V.: source of variance; DF: degree of freedom. CV: coefficient of variance; GP: germination percentage; SL: shoot length: RL: root length; SFW: shoot fresh weight; RFW: root fresh weight; SRR: shoot length/root length ratio; GR: germination reduction; SLR: shoot length reduction; RLR: root length reduction; GSTI: germination stress tolerance index; SLSTI: shoot length stress tolerance index; RLSTI: root length stress tolerance index. The final germination percentage, evaluated on the seventh day after salt stress initiation, was strongly affected by the stress level, thus showing decreasing trends as NaCl increased (GP: 86.35%, 23.33%, 15.00% and 8.40% at 0 mM, 100 mM, 200 mM and 300 mM NaCl, respectively) ( Table 2). The analysis of data further revealed significant differences in shoot (SL) and root length (RL) among the stress levels applied, with most profound effects being noted at the high stress level (Table 2). Accordingly, the shoot/root ratio (SRR) was considerably affected by the stress level, presenting a gradual increase as NaCl concentration increased (ranging from 1.28 to 3.58), which reflects the fact that root length was more profoundly affected than shoot length under stress conditions (Table 2). Such drastic effects were also noted in shoot (SFW) and root fresh weight (RFW), which were significantly affected by the stress level. In particular, both SFW and RFW showed decreasing trends as NaCl concentration increased, with the respective values ranging between 0.90-0.31 g and 0.14-0.04 g for shoots and roots ( Table 2).
In addition, a significant effect of the salt stress level on the germination stress tolerance index (GSTI) was also recorded, with the effects being proportional to the stress level applied. In particular, GSTI decreased as NaCl concentration increased (ranging from 0.29% to 0.09%), thus reflecting the differential effects of the various stress levels on seed germination potential (Table 2). Accordingly, the shoot length tolerance index and root length stress tolerance index (SLSTI and RLSTI) declined with increasing stress level. The effects of stress on root length were profoundly evidenced at higher stress levels where significant differences between the levels of 200 mM and 300 mM of NaCl were recorded, whereas in the case of shoot length, no significant differences in SLSTI values were recorded between the indicated stress levels (e.g., 200 mM and 300 mM of NaCl). In agreement with the abovementioned data for shoot and root fresh weight, the most drastic effects were noted at RLSTI compared to SLSTI, as a result of the most severe effects of salt stress on seedling root length, which are expressed with lower values of the RLSTI index (Table 2).

Effect of Landrace on Traits Related to Seed Germination and Seedling Growth Potential
The data underline the significant effect of landrace (p < 0.001) on the germination and seedling growth potential of squash under salinity stress conditions (Tables 1 and 3). The differential response of the landraces to salt stress was evidenced for all the traits under study, while the analysis showed RFW, SRR, SLR, RRL and GSTI as the most variable traits (CV > 20%) ( Table 1).
Germination was considerably affected by the landrace, as evidenced by the mean GP of landraces under study (Table 3), which ranged from 49.33% to 16.75% in landraces "747" and "1008", respectively. Although such values suggest a high salt tolerance ability and sensitivity for "747" and "1008" landraces, respectively, they are also attributed to their innate germination potential, as evidenced by the differences in GP in the respective control treatments (97.33% and 67.00% for "747" and "1008" respectively) (Table 4). Moreover, the landraces "746", "752", "745", "748", "749" and "750" were characterized by a mean GP of 39.50%, while the other landraces presented intermediate GP values ranging from 31.25% to 21.50% (Table 3). In relation to seedling growth potential, the effects of stress were consistently more severe on roots than on shoots (Table 3). Although the stress effects were obvious in all landraces, their response to salinity differed substantially. As such, "1006" presented the highest values for both SL and RL, while the lowest respective values were noted in "1007" and "1008". Regarding the SRR, "753" and "1007" presented the highest and lowest values, respectively. Accordingly, significant differences were also observed for SFW and RFW, with "752" showing the highest value for both traits, while the lowest values were recorded in "1008" and "746" landrace, respectively (Table 3). Table 3. The mean effect of the squash germplasm accessions on traits related to seed germination and seedling growth (mean ± SD), regardless of the salinity level (NaCl concentration). Means in the same column followed by the same letter are not significantly different at p < 0.05, according to Duncan's multiple range test; GP: germination percentage; SL: shoot length: RL: root length; SFW: shoot fresh weight; RFW: root fresh weight; SRR: shoot length/root length ratio; GR: germination reduction; SLR: shoot length reduction; RLR: root length reduction; GSTI: germination stress tolerance index; SLSTI: shoot length stress tolerance index; RLSTI: root length stress tolerance index. Table 4. Response of squash germplasm accessions to varying salt stress levels (0, 100, 200 and 300 mM NaCl) in relation to traits (means ± SD) related to seed germination and seedling growth.

300
"745" In agreement with the abovementioned findings, a considerable variation for the GSTI, SLSTI and RLSTI was also recorded (Table 3). Based on GSTI, "747" followed by "746" were proved to be the best performing landraces. In relation to SLSTI, "747" followed by "1006", "753" and "746" showed the highest tolerance to salt stress, whereas "1006" ranked as the best performing landrace in terms of RLSTI. In contrast, "1007", "758" and "759" were proved to be the most salt-sensitive based on GSTI, SLSTI and RLSTI values (Table 3).

Effect of the Landrace and the Salt Stress Level on Traits Related to Seed Germination and Seedling Growth Potential
Based on the analysis of variance applied on individual data, all traits related to germination and seedling growth potential under salt stress conditions were differentially affected by the landrace and the salt stress level applied (Table 4). In the absence of stress, germination was considerably affected by the landrace, thus substantiating a variable germination potential per se, which could be mainly attributed to the median longevity of seeds whose fruits were harvested at different periods (2018 to 2020). Among landraces, "746" and "1005" presented the highest and lowest GP under normal conditions (98.66 % and 63.00 %, respectively) which also justifies that seed longevity is a genotype-dependent variable (Table 4). Upon stress, the germination of all landraces was severely affected, with the effects of stress being in general analogous to its level. At 100 mM NaCl, "745" proved to be the most capable landrace of retaining a high germination rate (50.00%), while "746", "747" and "749" also showed relatively high GP values (≥40%). In contrast, "1007", "1008" and "1009" landraces, which were of medium-high innate germinability (67.00-96.00%), were incapable of germination at all stress levels (Table 4). Moreover, at 200 mM NaCl, "746" and "747" presented the highest GP (35.00 %), whereas "754" and "1006" ranked as the landraces with the lowest GP values. Finally, at 300 mM NaCl, "747" followed by "746" proved to be the best performing landraces (25% and 20%, respectively). Interestingly, "745", although showing superior performance at 100 mM NaCl, suffered great losses at both 200 mM and 300 mM NaCl (15.00% and 5.00%, respectively). Furthermore, the response of most landraces to salt stress involved a drastic reduction in both SL and RL, with the latter parameter being more pronouncedly affected in the majority of the tested landraces (Table 4). At all stress levels, "1006" presented the highest values for SL and RL, thus proving its ability to withstand even high levels of salt stress. In contrast, "750" consistently showed low values for SL and RL-a finding that could be partly attributed to its low growth potential per se, as evidenced by the respective values in the control treatments (6.83 and 5.10 mm, respectively) ( Table 4). In agreement with these findings, the SFW and RFW followed a decreasing trend as NaCl concentration increased. The highest values for these traits were recorded in "752" both in control and stress conditions, whereas the lowest values were noted in "1006" and "746" for SFW and in "747" for RFW. Although affected at all stress levels, the SRR was more profoundly affected at 300 mM NaCl, with the landraces showing varying values ranging from 10.19 to 1.27. The highest SRR was noted in "753", thus reflecting the more drastic effect of salt stress to the roots than in shoots ( Table 4).
The seeds of the selected landraces representing the four types of cultivated squash are illustrated in Figure 1. Regarding the different salt concentrations, landrace "746" (Batati orange) showed the highest germination percentage at all salt levels and was distinguished for its good tolerance to salt stress. Moreover, landrace "751" (Bejaoui green) expressed the lowest values especially at 200 and 300 mM, while landrace "748" (Kerkoubi orange) and "747" (Galaoui) could be also considered as similarly tolerant landraces and would be recommended for cultivation in saline soils or with irrigation water with a high salt level.

Effect of the Salt Stress Level on the Content of Malondialdehyde (MDA), Free Proline and Chlorophyll a and b
The results indicate that the different stress levels significantly affected the content of MDA, free proline and chlorophylls of squash seedlings (Table 5). Regarding MDA content, a significantly increased content was noted in stressed seedlings as compared to the control treatment (1.16 µmol mg −1 FW). The highest MDA content was recorded at 100 mM NaCl (1.75 µmol mg −1 FW), while the respective values for 200 and 300 mM NaCl did not differ significantly (1.43 and 1.34 µmol mg −1 FW). The content of free proline was the highest at the highest salinity level (0.92 µg mg −1 FW), whereas no significant differences were observed between the control and the 200 mM NaCl treatments (Table 5). An exception to this trend was noted at 100 mM NaCl, which showed the lowest values of free proline content (0.53 µg mg −1 FW). As expected, chlorophyll a (chla) and b (chlb) were reduced in stressed plants, as compared to the control treatment (Table 5), although the salt level of 100 mM NaCl showed a differential trend of increased chla and chlb   The obtained results point to a significant effect of the landrace on the content of MDA, free proline and chlorophyll in squash seedlings (Table 6). In relation to MDA, the highest content was recorded in landrace "746" (1.60 µmol g −1 FW), while the values of all the other landraces did not differ significantly. The landrace "746" was further distinguished by the highest content of free proline (1.00 µg mg −1 FW). On the other hand, landrace "748" showed the lowest content of free proline (Table 6). With respect to chla and chlb, the lowest values were obtained in the landrace "746" (12.49 and 25.69 mg g −1 FW of chla and chlb, respectively), followed by "747" (14.81 and 27.89 mg g −1 FW of chla and chlb, respectively), whereas the highest content was noted in "751" (51.07 and 108.74 mg g −1 FW of chla and chlb, respectively) ( Table 6). Table 6. Mean effect of the squash germplasm accession on the content of malondialdehyde (MDA), free proline and chlorophyll a and b (mean ± SD), regardless of the salinity level (NaCl concentration).

Effect of Landrace and Salt Stress Level on the Content of Malondialdehyde, Free Proline and Chlorophyll a and b
Our findings indicate that the contents of MDA, proline and chlorophyll were differentially affected by the landrace and the salt stress level applied (Table 7). In relation to MDA, "748" landrace showed a decreasing trend as NaCl increased. Furthermore, all the other landraces exhibited a decreased content upon stress, yet they showed a notable increase either at 100 mM NaCl ("747" and "746") or 300 mM NaCl ("751"). Accordingly, the response of these four landraces ("747", "746", "748" and "751") to the different stress levels varied considerably in terms of the free proline content ( Table 7). The stress response of most landraces involved an increased content of free proline, with the exception of landrace "746", whose content was maximized in control plants. In relation to the stress level effects, the application of 200 mM NaCl induced proline accumulation in "748" and "751", whereas "747" presented a peak at 300 mM NaCl (Table 7). Similarly, the content of chla and chlb was differentially affected both by the landrace and the stress level. Chlorophyll a and b were generally reduced in stressed plants, as compared to the control treatment, yet deviations from such a decreasing trend were observed depending on the landrace and stress intensity (Table 7). Specifically, "748" and "751" landraces showed the highest content of chla and chlb at 100 mM NaCl, while "747" presented the highest values at 300 mM NaCl. For the "746" landrace, the highest and lowest values for chla and chlb were noted at 0 mM and 200 mM NaCl, respectively. Table 7. Response of squash germplasm accession to varying salt stress levels (0, 100, 200 and 300 mM NaCl) in relation to the content of malondialdehyde (MDA), free proline and chlorophyll a and b (mean ± SD).

Discussion
The efficient selection of a salt-tolerant germplasm is considered of utmost importance in all breeding programs aimed at the development of salt-tolerant varieties. Given that salinity poses severe constraints to crop growth and productivity as well as to fruit quality in squash, this study aimed at determining the response of Tunisian squash germplasm to salinity stress at the stage of germination and early seedling growth. Plant response was assessed on the basis of selected traits related to seed germination and seedling growth potential. Based on their performance, four landraces were selected and further evaluated at the biochemical level, focusing on the determination of compounds that play a key role in the ability to withstand salt stress (e.g., MDA, free proline and chlorophylls).
Our findings revealed that salinity stress substantially affects all traits associated with germination and early seedling growth, with the effects of stress being in most cases analogous to the stress level applied. Differences in germination potential were reflected at the final germination percentage, on the seventh day after salt stress initiation, as well as at the decrease in germination over the control treatment. These findings are consistent with previous reports regarding the effects of salt stress on the germination of various plant species, including lettuce [33,40], rice [41], chickpea [31] and wheat [42]. Indeed, plants' sensitivity to salinity varies among growth stages [21], while during germination, the stress effects are aggravated due to the reduced seed water uptake that limits imbibition and seed turgescence [25]. Our data point to a gradual increasing severity of effects depending on stress intensity, thus supporting previous conclusions that the varying stress levels differently affect germination and seedling growth in various species, such as sugar beet and cabbage [43], soybean [44] and lentil [32]. Although all landraces were drastically affected, especially at high stress levels, their stress response varied significantly, thus indicating the existence of considerable genetic variation related to salt tolerance in the germplasm under study. As such, "746" and "747" were the best performing cultivars across stress levels, whereas "1007", "1008" and "1009" landraces, despite their relatively high innate germinability, were incapable of germination at all stress levels.
In addition to seed germination, the increasing level of salinity led to a gradually decreasing tissue elongation, expressed as reduced root and shoot lengths, probably as a result of toxicity as well as limited nutrient and water uptake due to osmotic stress [45,46]. Such data further support previous evidence that root and shoot lengths are the most suitable traits in terms of evaluating salt tolerance [31,47], since roots are responsible for water absorption and shoots for supplying aboveground tissues with water. In our study, seedling growth was inhibited at all salt stress levels, while root length was more severely affected than shoot length, most probably due to the fact that roots are directly and for a longer period exposed to salinity [46]. Indeed, such restricting effects have been attributed to the reduced water and nutrient absorption capacity from soil solution, which limits cell elongation and plant tissue development [48][49][50][51]. These findings are in agreement with the observed severe effects of salinity in roots of sugar beet, cabbage and amaranth [43] as well as soybean [44] and lentil [32]. Despite the fact that salinity led to a drastic reduction in seedling growth in all the tested landraces, a large variation between landraces was observed. At all stress levels studied, "1006" showed the highest SL and RL, therefore providing evidence for its salt tolerance ability, whereas the lowest values were noted in landraces "1007" and "1008" (in the case of SL) or "1008" (in the case of RL), indicating high susceptibility to salt stress. Accordingly, our data support a trend of gradually decreasing seedling fresh weight with increasing stress intensity, which could be mainly attributed to ionic effects occurring as a result of a proportional increase in Na+ concentration. Among landraces, "752" showed the highest growth potential as expressed by SFW and RFW parameters, both under normal and salt stress conditions, whereas "1008" and "747", "1008" and "1009" were the most drastically affected in relation to SFW and RFW, respectively. Concomitant with the abovementioned findings, all the recorded physiological indices were severely affected by salinity stress, especially at high stress levels. The adverse effect of salt stress on GSTI, SLSTI and RLSTI has been previously evidenced in maize germplasm subjected to drought stress [52], while previous reports further underline the significant effect of the genotype on the response to salt stress for several crop species [43,51,53]. Our findings support the notion that stress effects are differentially expressed in the germplasm under study. As such, "747" proved to be the most superior landrace in terms of GSTI and SLSTI, while "1006" was characterized as the most tolerant landrace based on RLSTI. In contrast, the combined data of GSTI, SLSTI and RLSTI classified "1007", "1008" and "1009" as the most salt-sensitive landraces.
In addition to germination and seedling growth potential under stress conditions, an indicator routinely employed to assess the degree of damage to plant cells caused by abiotic stress is MDA accumulation-a natural product of lipid peroxidation due to oxidative stress [54][55][56]. Our data revealed a significant variation both in the landraces tested and the stress level applied. Although the mean MDA content across landraces was generally increased in stressed plants, as compared to the control treatment, such an increase was attributed to a peak in MDA content at specific stress levels. Interestingly, "746" and "747" showed the maximum content at 100 mM NaCl, while in landrace "751", MDA content reached its peak at 300 mM NaCl. Such findings are in total agreement with the previously reported inconclusive differences in MDA level in purslane plants subjected to varying salt stress levels, where the MDA content showed a peak after exposure at 300 mM NaCl for 6 or 9 days [57]. In general, the increased MDA and O2-synthesis under salt stress has been associated with damage to cell membrane integrity and protein activity, although the damage recorded depends on the stress period [58]. It has been suggested that long-term exposure to high salinity may lead to the destruction of cell membranes, adversely affecting SOD, POD and CAT activities, whereas oxidative stress-mediated lipid peroxidation does not occur within a short period; e.g., up to 5 week exposure to salt stress [59]. Our data further point to inconclusive differences in free proline content at various stress levels. Although the response of most landraces to salt stress involved an increase in free proline, with the exception of "746", the maximum content was observed either at 200 mM ("748" and "751") or at 300 mM NaCl ("747"). Such observations further support the previously reported negative correlation between MDA and free proline content [54], yet they partly deviate from the suggested positive association between proline content and stress intensity [57]. In our study, this association was only evidenced in landrace "747", whose overall performance classified it as salt-tolerant, thus supporting previous conclusions related to the increased proline content in salt-tolerant genotypes of various plant species, including potato [60], melon [61] and tomato [62]. It is well evidenced that plants' response to high salinity adversely affects photosynthesis activity, thus involving a reduction in chlorophyll content. Our data showed a general decreasing trend of chla and chlb in stressed plants, as compared to the control treatment, which indicates that the response to high salinity may involve a decrease in chlorophyll content in various plant species, including pepper [63] and winter squash [64]. However, deviations from this trend were noted among the tested landraces and the stress levels applied, indicating a genotype-depended response. Interestingly, the content of both chla and chlb was the maximum at 100 mM NaCl across landraces, with this peak being mostly evidenced in landraces "748" and "751", thus probably reflecting an adaptive stress response governing tolerance to low-medium salinity.
Overall findings provide conclusive evidence that the evaluation of squash genotypes at germination phase shows a great potential for revealing genetic variability related to salt tolerance. Addressing the classification of genotypes in terms of salt tolerance, our data related to germination and seedling growth potential under salt stress point to the superiority of landraces "746" and "747", followed by "1006", at all stress levels applied. In contrast, "1007", "1008" and "1009" landraces were incapable of germination at all stress levels, thus proving their sensitivity to salinity even at relatively low stress levels. Our findings further support previous reports highlighting that the innate genotypic high growth potential is associated with salt sensitivity and vice versa [65]. Mostly relevant to this argument is the performance of landrace "751": although it exhibited an enhanced seedling growth under normal conditions, it suffered a significant decrease after exposure to salt stress. Addressing the evaluation of salt tolerance based on biochemical parameters, the observed cumulative patterns for MDA, free proline and chla and chlb deviate considerably from the general stress response profiles that usually involve an increased MDA and proline content as well as a decreasing chlorophyll content with increasing salt concentration. Despite such deviations, which further strengthen previous reports on inconclusive differences in MDA and chlorophyll content in various plant species subjected to salt stress [57,65,66], the overall data support the feasibility of conducting the early selection of salt-tolerant squash germplasm on the basis of germination and seedling growth potential under salt stress.

Plant Material and Growth Conditions
The studied genetic material consisted of 15 Tunisian squash landraces, collected from different geographic regions of Tunisia during the period extending from 2018 to 2020 (Table 8). Each landrace was assigned passport data and an inventory number, according to the National Gene Bank of Tunisia, while full details are available at the Germplasm Resources Information Network-GRIN (http://www.tn-grin.nat.tn/gringlobal/search. aspx, accessed on 15 February 2022). The description of fruit morphology was performed based on the European Cooperative Program for Plant Genetic Resources (ECPGR) list of descriptors for Cucurbita spp. [67].

Salinity Stress Treatments
The experiment was carried out at the Department of Horticulture, High Agronomic Institute of Chott Mariem-Sousse-Tunisia. Following the selection of seeds for size homogeneity, 50 seeds per landrace (five petri dishes with 10 seeds each, n = 5) were surfacesterilized for 5 min in 10% H 2 O 2 (v/v) and rinsed twice in sterile dH 2 O. Sterilized seeds were primed via exposure to an eliciting solution of 1.5 mM gibberellic acid (GA 3 ) for 24 h to stimulate germination and subsequently rinsed in sterile dH 2 O. Five to ten seeds, according to size, were placed on sterile petri dishes containing two layers of filter paper moistened daily with 5 mL of appropriate solutions: dH 2 O (control), 100, 200 and 300 mM NaCl. Seedlings were grown under controlled conditions for 7 days (25 ± 2 • C, 50 ± 5% relative humidity, 18 h light/6 h dark photoperiod under white fluorescent light (40 µmol m −2 s −1 ).

Determination of Germination and Seedling Growth Potential under Salt Stress
Salt tolerance was evaluated on the basis of various parameters related to seed germination and seedling growth potential under salt stress conditions, measured daily until no more germinated seeds were recorded (first-seventh day) ( Table 9). Seeds were considered germinated when the protruding radicle was at least 2 mm long. The parameters germination reduction (GR), root length reduction (RLR) and shoot length stress tolerance index (SLSTI) express the decreased values of salt-stressed plants over the control treatment.

Evaluation of Salinity Tolerance Based on Biochemical Parameters
Based on the obtained data related to the response of the 15 landraces to salinity stress, 4 landraces representing the main types of cultivated squash were selected for further evaluation. As such, the landraces "748" (Karkoubi), "751" (Bejaoui), "747" (Galaoui) and "746" (Batati) were assessed, employing the content of seedling tissues in MDA (MDA), free proline and chlorophyll a and b as evaluation criteria.
The content of MDA (MDA) was determined using the method applied by Hnilickova et al. [57] with minor modifications. Briefly, 200 mg of leaf samples were homogenized with liquid nitrogen and, following the addition of 80% ethanol, samples were centrifuged for 5 min at 6000 rpm. Aliquots of 0.7 mL of the supernatant solution were mixed with 0.7 mL of 0.65% thiobarbituric acid (TBA) in 20% TCA (trichloroacetic acid) and 0.01% BHT (butylated hydroxytoluene). A second set of 0.7 mL samples was mixed with 0.7 mL of 20 % TCA and 0.01 % BHT. Following incubation at 95 • C for 25 min and subsequent cooling, samples were centrifuged for 5 min at 6000 rpm. The content of MDA was determined at 532 nm using a UV-Vis spectrophotometer (Evolution 210, Thermo Scientific, Abingdon, UK) and expressed in µmol g −1 of fresh weight (FW).
The content of free proline was measured according to the method described by Monneveux and Nemmar [68]. Leaf samples (100 mg) were homogenized in 10 mL of 3% sulfosalicylic acid and, following filtration, the homogenate was heated to 85 • C in a water bath for 60 min. After cooling, 1 mL of ninhydrin reagent was added (ninhydrin reagent consisted of 120 mL distilled water, 300 mL of acetic acid, 80 mL acetic orthophosphoric acid at a density of 1.7, and 25 mg of ninhydrin). The samples were boiled for 30 min and, after cooling, 5 mL of toluene was added and samples were vortexed. The upper phase was recovered and was measured using a UV-Vis spectrophotometer (Evolution 210, Thermo Scientific, Abingdon, UK) at 528 nm. A proline standard curve ranging from 0 to 2.5 mg mL −1 of L-proline was used to determine the proline content, expressed in µg mg −1 of FW.
For chlorophyll content determination, the extraction of samples was performed as described by Curtis and Shetty [69]. Briefly, 50 mg of leaf tissue (in triplicate) was extracted into 3 mL of methanol and stored at 23 • C in darkness for 2 h. Absorption of extracts (1.5 mL) was measured at 650 and 665 nm using a spectrophotometer (Evolution 210, Thermo Scientific, Abingdon, UK). Chlorophyll a and chlorophyll b were expressed in mg g −1 FW.

Statistical Analysis
The experimental layout was completely randomized with three replications. Data were analyzed using ANOVA tests (p ≤ 0.05), according to the experimental design, combining salt concentrations and genotypes. Differences between means were compared using the Duncan Multiple Range test (DMRT). Statistical analyses were performed using SAS software V9 (SAS Institute, Cary, CA, USA).

Conclusions
Soil salinization is gradually increasing due to the scarcity of rains and the increase in evapotranspiration, adversely affecting plant germination, growth, development and fruit setting in salt-affected soils, as is the case of semiarid regions. Most vegetable species are salt-sensitive at all stages of their lifecycle, while germination and seedling growth have been viewed as the critical stage under salt conditions. In particular, salinity is a factor that severely limits squash growth and productivity while, at the same time, deteriorating fruit quality. Despite the scientific evidence in relation to the salt stress response in a plethora of plant species, there is a lack of information in relevant research fields for squash. The development of a salt-tolerant germplasm is one of the most effective means for enhancing squash production in saline soils. The results of our study suggest the feasibility of conducting an early selection of salt-tolerant squash germplasm on the basis of germination and seedling growth potential under salt stress. Such an approach may considerably upgrade all procedures aimed at selecting salt-tolerant germplasm to be exploited for cultivation in saline soils. In this context, landraces "746" and "747" were the best performing cultivars for the tested salinity levels in terms of germination percentage, germination percentage stress index and shoot length stress index, which indicates that they could be integrated as valuable germplasm material in breeding programs targeted at improving salt tolerance in squash through the selection of elite genotypes.