Salinity represents one of the major limitations for yield and quality of a number of different crops (Maggio et al., 2011). According to the estimates of the Food and Agriculture Organization of the United Nations, about 20% of irrigated land worldwide is affected by the increase of the salinity level (Rozema and Flowers, 2008). This phenomenon is accentuated by the competition for fresh water between agricultural and civil uses, which is worsened by climate changes, growing population (Maggio et al., 2011), socio-economic development and water contamination (Balia and Viezzoli, 2015). Salinity conditions are relevant not only to arid and semiarid environments and to the southern regions of the world, but also to the Mediterranean coastal areas. In Europe, 26 countries (Maggio et al., 2011), in particular Spain, Portugal, Italy, Greece (Ghiglieri et al., 2012) and France (Puard et al., 1999) are interested by salinization phenomena (Maggio et al., 2011). In the coastal areas, salinization of aquifers is usually caused by saltwater intrusion (Mongelli et al., 2013) as a result of groundwater overexploitation (Balia and Viezzoli, 2015).
In Italy, this phenomenon is found in various regions, such as Sardinia (Capaccioni et al., 2005), the Catania Plain (Capaccioni et al., 2005), Tuscany (Barrocu, 2003), the Tiber Delta (De Luca et al., 2005), Campania, Calabria (Barrocu, 2003) and the Adriatic coast (Ghiglieri et al., 2012), especially the Po Plain (Antonellini et al., 2008).
Some of these areas, in particular the Po Plain and the Oristano province (Sardinia), are used for rice cultivation. In these areas, the process of salinization may contribute to reduce crop yield, together with the presence of highly problematic weeds. In fact, salinity represents one of the environmental conditions that affect seed germination and plant growth (Sadeghloo et al., 2013), both in weeds and in crops. The knowledge of the ability of seeds to germinate in different environmental conditions, including salinity, is considered of fundamental importance not only for crop establishment, but also for estimating weed development in agricultural ecosystems (Koger et al., 2004; Benvenuti, 2011).
Among the main weed species that infest rice fields, barnyard grasses (several species of the genus Echinochloa P. Beauv.), weedy rice (Oryza sativa L.), and sedges (mainly the genus Cyperus L.) are some of the most troublesome, able to cause significant yield losses (Panozzo et al., 2013).
In particular, Echinochloa spp. have been key weeds in almost all rice systems worldwide, including European and Italian rice fields. These weeds are characterized by C4 photosynthetic pathway and some species are able to grow both in dry and flooded conditions (Vidotto et al., 2007). Echinochloa spp. exhibit great competition effects towards rice, especially during early stages of cultivation. Gibson et al. (2002), for example, found that the competition established by Echinochloa spp. is significantly lower if a rice field is maintained free from these weeds during the first 30 days after seeding.
Echinochloa spp. can be distinguished in red or white biotypes on the basis of the different pigmentation at the basal sheaths of the plant. It has been established that this different plant pigmentation could reflect differences in herbicide sensitivity of the weed (Tabacchi et al., 2006). Problems associated with Echinochloa spp. management have worsened in the last decade, due to the selection of populations resistant to different herbicides applied in rice field, in particular ALS and ACC-ase inhibitors (Panozzo et al., 2013). In the search of alternative and effective techniques for controlling these weeds, several methods have been tested, including biocontrol agents (Hershenhorn et al., 2016) and the use of rice varieties tolerant to herbicides (Kraehmer et al., 2017).
Even though the area of rice cultivation affected by salinization is increasing at global level, the response of rice weeds to salt conditions and the potential influences on weed ecology and competition effects have been poorly investigated so far. The stress due to the presence of salt causes physiological and biochemical alterations and tolerance to salinity in plants is related to the synthesis (induced by the stress) of several compounds, including abscisic acid (ABA), glycinebetaine, organic acids, proline, polyamine and polyols (Cowan et al., 1992). The functions of these compounds in response to salt stress seem to be those of chemical signals, osmotic adjustment, free radical scavenging, preservation of proteins and membrane integrity. As concerns the grass weeds, it was found that the tolerance to saline conditions is due to their capacity in uptake and translocation of Na+, maintaining K+, and osmotic regulation through the accumulation of proline and glycinebetaine, even though salt tolerance in E. crus-galli seems to be not related to Na+ adsorption changes (Yamamoto et al., 2003).
In the case of Echinochloa spp., in particular, it is not yet entirely clear what is the effect of salt stress on their behaviour and physiological pathways. In these terms, a potential delay in germination or a negative influence on seedling establishment and first growth due to salinity could potentially alter the role of these weeds in rice production systems. In addition, the knowledge of the interactions between response to salinity and other agronomical relevant traits, such as herbicide resistance, is essential to estimate potential future impacts of salinity on Echinochloa spp. management in rice. In particular, it is not clear if there is interaction between herbicide resistance and salt stress and what could be the biochemical mechanisms involved.
This study aimed to evaluate the effect of salinity on germination of different common barnyard grass (E. crus-galli (L.) P. Beauv.) populations collected in the Italian rice area and to verify the presence of differences in terms of salt sensitivity between populations that are susceptible or resistant to ALS-inhibitor herbicides.
Materials and methods
Seeds of six common barnyard grass populations (E. crusgalli) were used during the trials. The seeds were collected between 2010 and 2012 in Italian rice fields, which underwent repeated applications of penoxsulam since many years. Penoxsulam is a broad-spectrum ALS-inhibitor herbicide used in rice to control common barnyard grass, water-plantain (Alisma plantago-aquatica L.), red stem (Ammannia coccinea Rottb.) and other weeds. After collection, the seeds were let dry at room temperature for about one week and then stored at +5°C. At different timings, seedlings from an aliquot of the collected seed lots were tested for resistance to ALS-inhibitor herbicides via a greenhouse bioassay and a molecular study that confirmed the presence of target site resistance in three populations (labelled r1, r2, and r3) due to Trp-574-Leuc point mutation in the ALS gene (unpublished data). Thus, the other three populations (s1, s2, and s3) were considered sensitive to penoxsulam. For the present study, seeds of the six populations were taken for the same collected seed lots.
Effect of water salinity on germination
In order to evaluate the effect of water salinity on the germination, 20 seeds for each population were placed in Petri dishes (9 cm diameter) lined with one filter paper imbibed with 5 mL of deionized water or saline solution.
Nine different salt concentrations were applied. Salt solutions were prepared by dissolving NaCl in deionized water at the following concentrations: 0 mM, 50 mM, 100 mM, 150 mM, 200 mM, 250 mM, 300 mM, 350 mM, and 400 mM. These nine salt concentrations were selected among those applied in previous studies (Chauhan and Johnson, 2009; Sadeghloo et al., 2013; Opena et al., 2014) and according to the water salinity levels found in some European rice cultivation areas (Isla et al., 2003; Gay et al., 2010; Moret-Fernandez and Herrero, 2015) in order to be able to compare the results of our trial with those found in other studies carried out in analogous conditions and in order to use saline concentrations that can be found in the area from which the six common barnyard grass populations tested came from. The number of doses was selected following the guidelines provided by Holland-Letz and Kopp-Schneider (2015).
Three replicates for each salt treatment were used. Petri dishes were sealed with Parafilm to avoid drying and contamination. Afterward, they were incubated in a growth chamber at a constant temperature of 25°C and arranged in a randomized complete block design.
Seed germination was recorded every day for 15 days. Moreover, at the end of the incubation period (15th day), length of roots and shoots of a sample of 10 seeds for each Petri dish was determined.
Two runs of the entire experiment were carried out.
ANOVA was carried out on data of germination capacity (total of number of seeds germinated by the end of germination test), speed of germination and length of roots and shoots in order to test the effects of population, NaCl concentration and experiment run. The test was conducted using the open source programme and environment R (R Core Team, 2016)
A regression analysis was performed on data of germination capacity. The fitted model was the following 3-parameter loglogistic regression model (Streibig et al., 1993; Knezevic et al., 2007; Vidotto et al., 2013):
where Y is the germination capacity, x is the NaCl concentration in mM, d is the upper limit, and b is the relative slope at the point of inflection e.
Model fitting was performed using the function drm of the add-on package drc of the programme R (Ritz and Streibig, 2012; Ritz et al., 2015).
The same model was also used to describe the effect on speed of germination, as the relationship between duration of germination test and cumulated number of germinated seeds. The analysis was conducted separately for each salt concentration with the time (days) as the parameter x in Eq. (1) and cumulated germination as the dependent variable Y. Furthermore, the same analysis was applied to model the relationship between NaCl concentrations (independent variable) and either the length of roots or length of shoots (dependent variables).
The effective concentrations required to reduce by 50% either germination capacity or shoot or root length in comparison to the values obtained at 0 mM salt concentration (EC50) were estimated from the fitted models using the function ED of the package drc. In the case of cumulated germination, the number of days required to obtain 50% of final total cumulated germination was calculated. The values of EC50 were used to perform pair-wise comparisons between populations, or between averages among resistant or susceptible populations, by calculating a Sensitivity Index (SI):
where A and B refer to two generic populations under comparison. SI was also calculated by considering EC50(A) and EC50(B) in Eq. 2 as estimated by pooling data of resistant or susceptible populations, respectively.
The significance of SI of each comparison was calculated by using the function EDcomp of the package drc of the R programme.
Results and discussion
The results showed that the germination of the tested weed biotypes was affected by the salt treatments (Figure 1).
The analysis of variance showed significant differences in terms of germination capacity between resistant and sensitive populations from 0 mM NaCl to 250 mM NaCl (Table 1). At these saline doses, the germination capacity of the sensitive populations ranged from about 84% to 94% and resulted greater than that of the resistant ones, which ranged from about 60% to 71%.
Seed germination remained quite stable up to a saline concentration of 250 mM, with values slightly above 60% and 80% for resistant and sensitive populations, respectively. These germination values were similar to those recorded in previous studies on E. glabrescens Munro ex Hook in which at a salt concentration of 200 mM the germination was more than 60% and averaged about 73% at 250 mM NaCl (Opena et al., 2014); similar germination level (68%) was also found on E. crus-galli at a salt concentration of 225 mM (Sadeghloo et al., 2013). Conversely, in trials carried out by Chauhan and Johnson (2009), germination of E. colona (L.) Link was totally inhibited at the salt concentration of 200 mM. In our study, increasing salt concentration above 250 mM reduced seed germination, until about 13% germination at 400 mM NaCl, with no differences between resistant and sensitive populations. At 400 mM NaCl germination was inhibited for the populations s2 and r2, while it ranged from 5% to 33% for the populations s1 and s3, respectively (Figure 1). At the same NaCl concentration Sadeghloo et al. (2013) found that germination was completely inhibited. Previous studies on E. colona and E. glabrescens also found a linear decrease of germination with the increase of salt concentration (Chauhan and Johnson, 2009; Opena et al., 2014).
The EC50 for germination capacity was the lowest for the resistant population r2 (EC50 274.47 mM), while the highest for the sensitive population s3 (EC50 380.73 mM) (Table 2). The pairwise comparison between populations showed significant differences between population s2 and population r3 and between population r2 and s3 (Table 3). Furthermore, significant differences were also found between the resistant populations r2 and r3. According to these results, there is no a clear evidence that response to saline conditions was influenced by sensitivity towards ALS-inhibitor herbicides, as in some cases significant differences were found between populations showing a similar herbicide sensitivity (such in the case of s2 vs s3 and r2 vs r3).
Speed of germination
The results obtained showed that the appearance of the first germinated seeds delayed with the increase of NaCl concentration (Figure 2). An analogous trend was found by Hakim et al. (2011) during a germination test conducted on different weeds, including E. crus-galli and E. colona. At 0 mM NaCl concentration the seeds began to germinate between the second and the third day from the start of germination test and reached their maximum germination between the fourth and the ninth day. Furthermore, at 0 mM NaCl only populations s1 and r2 showed significant differences in the time required for achieving 50% of their final germination capacity: in particular, population s1 required 2.48 days while population r2 needed 3.84 days (Table 4; Figure 2).
For the resistant populations, the time required to reach 50% of germination capacity ranged, on average, from a minimum of 2.85 days at 50 mM NaCl to a maximum of 12.42 days at 400 mM NaCl; averaging among sensitive populations, EC50 ranged from a minimum of 3.42 days to a maximum of 10.26 days, respectively, at 50 mM and 400 mM salt concentration (Figure 2). At the higher saline concentrations, the barnyard grass populations tested in our trials showed a speed of germination greater than that found by Hakim et al. (2011): in fact, in their study, the authors observed that E. crus-galli and E. colona began to germinate after 12-15 days. Pairwise comparisons showed the presence of significant differences between resistant and sensitive populations at 50 mM, 200 mM, 250 mM and 300 mM NaCl concentrations (Table 4). Furthermore, significant differences were found within resistant and within sensitive populations at salt concentrations of 50 mM (s1 vs s2 and r2 vs r3), 200 mM (r2 vs r3), 250 mM (r1 vs r2) and 300 mM (s1 vs s2). As observed for germination capacity, also in the case of speed of germination there is no a clear evidence that sensitive and resistant populations react differently to increasing NaCl concentrations.
Length of roots and shoots
The results showed that, in all the tested populations, the length of roots decreased with increasing NaCl concentrations (Figure 3). Root length ranged, on average, from 9.88 cm at 0 mM NaCl to 0.36 cm at 400 mM NaCl. A similar trend was found by Hakim et al. (2011) on E. crus-galli and E. colona, in which a progressive reduction of root length with the increase of saline concentration was observed. The value of EC50 for root length in resistant populations averaged 162.60 mM, with a minimum of 96.36 mM for the population r2 and a maximum of 230.98 mM for the population r3. In the case of the sensitive populations, EC50 for root length averaged 159.76 mM, with a minimum of 114.81 mM for the population s2 and a maximum of 210.10 mM for the population s3 (Table 5).
The pairwise comparisons between populations underlined the presence of significant differences in root length among some of the resistant and sensitive populations (Table 6).
Regarding the length of shoots (Figure 4), a similar behaviour was observed, with shoot length decreasing with the increase of salt concentration. Shoot length ranged, on average, from 6.16 cm at 0 mM NaCl to 0.41 cm at 400 mM NaCl. Hakim et al. (2011) obtained analogous results on different species, including E. crusgalli and E. colona. In general, the values of EC50 for shoot length were higher than those recorded for root length, indicating that shoot growth was apparently less affected by salinity than root growth. In resistant populations, EC50 for shoot length was on average 246.10 mM, with a minimum of 226.47 mM for the population r2 and a maximum of 261.67 mM for the population r1 (Table 5). In the case of sensitive populations, EC50 for shoot length was 217.00 mM, with a minimum of 199.23 mM for the population s2 and a maximum of 249.67 mM for the population s3. The pairwise comparisons between populations underlined the presence of significant differences in shoot length among some of the resistant and sensitive populations (Table 6). As observed for germination capacity and speed of germinations, there are no clear trends of different tolerance to salinity conditions in the tested populations that could be attributable to sensitivity to ALS-herbicides. Nevertheless, a similar behaviour was observed in terms of response of shoot length to salinity within resistant populations.
The results obtained in this study suggest that in the tested E. crus-galli populations, seed germination capacity, speed of germination, root and shoot growth were affected by saline conditions.
A remarkable barnyard grass tolerance to moderate salinity levels was observed: from 0 mM to 250 mM NaCl the seed germination capacity was up to 90% in the sensitive populations while it was about 70% in the resistant ones. Moreover, one of the populations (s3) showed a good tolerance to salinity, with a percentage of germination equal to 33%, even at 400 mM NaCl. Although the resistant tested populations exhibited an intrinsically lower germination capacity, as shown by the lower values recorder in nonsaline conditions (0 mM NaCl), the response to salinity is similar to that observed in sensitive populations.
These results are interesting both from an ecological and agronomic perspective. For example, the reduction of speed of germination at increasing salt levels could suggest a reduced competitive activity of barnyard grass. Moreover, these results could be potentially exploited for predicting weed emergence dynamics through modelling, also in crops different from rice (Masin et al., 2010, 2012).
The real consequences in terms of competitions towards the crop should be evaluated also taking into consideration the negative impact that salinity could have also on germination and first growth of the crop itself. Future studies are then needed for assessing the response to salinity of the main rice varieties cultivated in the environment in which the E. crus-galli populations tested in this study were collected.