Mitigating thermal and water stress in lentils via cultivar selection and phosphorus fertilization

Human activity contributes to increasing the concentration of greenhouse gases in the atmosphere thereby increasing the radiation absorbed, and subsequently affecting the air temperature. Climate change is a daunting environmental challenge because of its direct biophysical effects on plant physiological processes and consequently on crop production. In recent years, extreme climate events, including heat waves, heavy precipitation and drought incidents, have strongly affected Europe and led to crop failure (Reidsma et al. 2010).

Although higher temperatures in northern Europe are expected to increase crop yields, in the Mediterranean and south-western Balkans increasing temperatures, in combination with reduced rainfall, are expected to lead to large yield reductions (Olesen et al. 2011). However, it is reported that elevated levels of carbon dioxide (eCO₂), due to climate change, can mitigate the effects of drought in grain legumes such as soybean (Li et al. 2013), as plants reduce stomatal and canopy conductance and therefore limit soil water absorption. While it is difficult to predict accurate scenarios regarding future climate change, crop productivity is a key factor affecting food security and economic stability (Hatfield et al. 2011).

Lentils (Lens culinaris Medik.) are grain legumes which have been grown around the Mediterranean basin since antiquity and still remain a structural block of the Mediterranean diet pyramid. Apart from their high nutritional value, legumes, and lentils in particular, constitute a food group which can effectively meet the demands imposed by climate change as the impact on the environment is minimal (Graham & Vance 2003). In the Mediterranean region, lentils are usually sown during the wet winter months, but decreasing soil moisture and increasing temperatures during the reproductive stages result in terminal drought and low seed yields (Shrestha et al. 2006). To cope with such phenomena, it is necessary to exploit genetic variation reported to reside among cultivated germplasm as well as within wild Lens spp. (Erskine et al. 2011).

Earliness in flowering is a desirable trait, especially under the semi-arid conditions prevailing in the Mediterranean region. Early sowing has been an effective strategy in many areas to avoid high temperatures and low rainfall during the reproductive stages, and thus to increase seed yields (Hajarpoor et al. 2014); however, earlier sowing was reported to increase susceptibility to disease due to longer favorable pre-winter weather conditions (Colbach et al. 1997). On the other hand, earliness owing to genotypic variation leading to drought avoidance could be exploitable. In addition, previous studies have reported the beneficial effect of increased phosphorus (P) rates on earliness in flowering and maturity. This effect is possibly caused by an alteration of the nitrogen/phosphorus (N/P) ratio in plant tissues and/or by the beneficial role of P in stimulation of the reproductive parts of plants (Sawan et al. 2008). Genotypic variation and environmental effects on water economy and productivity of C₃ species can easily and reliably be assessed by carbon isotope discrimination (Δ values) in above-ground biomass. Δ values (a measure of the ¹³C/¹²C ratio in plant tissues compared to the air) are related to the ratio of intercellular to ambient CO₂ concentrations (c_i/c_a) being a surrogate of the intrinsic water use efficiency [(WUE_i, the ratio of CO₂ assimilation rate to stomatal conductance (g_s)] (Farquhar & Richards 1984). Moreover, Δ values have been proved a reliable, indirect and long-term indicator of water use efficiency at biomass level (WUE, the ratio of biomass produced to the water consumed to produce it) for many species; relationships between Δ and yield have also been established (Turner 1996).

In the Mediterranean region, soils are usually P-deficient, thus the use of P fertilizers remains relatively high (Harmsen & El Mahmoud 2004). Phosphorus is an essential macronutrient with a pivotal role in numerous plant processes. Since lentils meets their needs for N through N₂-fixation, P is the most common limiting nutrient for both growth and grain yield (Yadav et al. 2009). Moreover, Jin et al. (2014) mentioned that eCO₂ increased drought tolerance as well as water use efficiency (WUE) of field pea (Pisum sativum), especially when plants are grown under sufficient P supply.

Temperature is perhaps the most critical factor in all biological processes, therefore heat units are often used to estimate phenological development of plant species. Growing degree-days (GDD) provide a simple estimate of the accumulated heat available during the growth season or life cycle of an organism (Robertson et al. 2013). The concept of degree-days is based on three assumptions: (i) there is a basal temperature for all species below which no growth is observed, (ii) the growth of all kinds is proportional to the total accumulation of energy over a specific period, and (iii) species' maturity occurs only when a specific total of degree-days is achieved (Fealy & Fealy 2008).

Therefore, the aim of this work was to study the effect of four P rates (0, 30, 60 and 90 kg P₂O₅ ha⁻¹) on the earliness of four lentil cultivars at seven growth stages, through the calculation of GDD. Furthermore, in order to elucidate whether P supply can affect WUE and seed yield under P-deficient Mediterranean soil, Δ values were determined in above-ground biomass at full bloom.

A field experiment was conducted for two growth seasons (2013–2014, hereafter 2014 and 2014–2015, hereafter 2015) in the farm of the Aristotle University of Thessaloniki (AUTh; 40° 32′12 N, 22° 59′21 E, 6 m a.s.l.). The climate is typically Mediterranean. Mean monthly temperature and monthly precipitation during the growth seasons (December to May) are shown in Figure 1 while Table 1 presents soil characteristics of the experimental site.

Under rainfed conditions, four lentil cultivars [Samos, Thessaly, Flip 2003-24 L (Flip), Ikaria], provided by the Hellenic Agriculture Organization (HAO)-DEMETER, Institute of Industrial and Fodder Plants, Larissa, Greece, were supplemented, before sowing, with four phosphorus (P) rates [0 (P₀), 30 (P₃₀), 60 (P₆₀) and 90 kg P ha⁻¹ (P₉₀)] in a split-plot design with P rates in the main plots and cultivars in the subplots. The treatments were triplicated. Phosphorus was applied as triple superphosphate (460 g P₂O₅ kg⁻¹). Seeding was conducted by hand on 6 December 2014 and 5 December 2015. Each subplot consisted of six rows, 4 m long and 0.25 m apart (6 m²), at a sowing density of 100–110 seeds m⁻². Breeder's seeds were used for all the four cultivars.

For each subplot, the Julian date, by which specific growth stages were achieved, was recorded. Seven growth stages, following Erskine et al. (1990), were monitored: two vegetative, the stages of 4–5th and 7–8th leaves (V4–5 and V7–8, respectively), and five reproductive stages, namely first bloom (R1), full bloom (R2), flat pod stage (R4), full pod filling stage (R6) and full maturity (R8). The achievement of the growth stage was considered when at least 50% of the plants in each subplot had reached the defined stage.

At full maturity (R8 growth stage), a 0.125 m² surface was harvested and threshed by hand in order to determine seed weight (SW) per subplot (in g m⁻²) after normalization at 10% seed moisture.

Data of GDD for each growth stage were subjected to analysis of variance (ANOVA) as an over-year, split-plot design with P rates in the main plots and cultivars in the subplots and with three replications. Means were compared by the least significant difference (LSD) test at P < 0.05 and analyses were run using the statistical software JMP 5.1 (SAS Institute Inc., Cary, NC, USA).

In the present study, over the years, cv. Flip had the highest SW (406.34 g m⁻²) while it was the earliest among the three cultivars; it was followed by Thessaly (343.03 g m⁻²), Samos (299.36 g m⁻²) and finally Ikaria (283.59 g m⁻²), the latest maturing cultivar. The effects of climate change include high temperatures and prolonged drought periods in the Mediterranean region (Iglesias & Garrote 2015); thus earliness in flowering is a desirable trait under the prevailing hot and dry conditions. Moreover, lentil, as a cool season legume, is more sensitive to high temperatures, especially when they occur during the flowering and pod-filling stages resulting in high yield losses (Farooq et al. 2017). In most lentil producing areas, yield does not exceed half the potential yield, mostly due to scarcity of water combined with extreme temperatures, indicating the need to increase crop yield through genetic improvement and/or better crop management (Ghanem et al. 2015).

Breeding for stress tolerant cultivars is more intricate because of the complex nature of the mechanism that gives tolerance. Thus, a range of different crop management practices can be applied to mitigate the effects of drought stress including sowing date, early maturing cultivars and appropriate fertilizer use. In the present work, neither cultivars nor P rates had a significant effect on GDD of the vegetative stages but both factors affected the earliness of the reproductive stages; the interactions P rate × cultivar and P rate × growth season were significant as well. More specific, at the beginning of flowering (R1 growth stage), cv. Flip was the earliest showing no significant differences between P rates. On the contrary, in cvs. Samos, Thessaly and Ikaria, P additions induced earliness compared to the control (P₀) treatments (Table 2). The highest GDD were recorded for P₀ in cvs. Ikaria and Samos (1033.30 °C and 1042.15 °C, respectively). Similar results were evident at full bloom (R2 growth stage) when cv. Flip recorded the lowest GDD (998.89 °C) regardless of the P rate. At flat pod stage (R4), P additions began to stand out in comparison to P₀. Treatments Ρ₆₀ and Ρ₉₀ induced earliness in Flip and Ikaria while in cv. Thessaly, all the P additions (P₃₀, P₆₀ and P₉₀) scored lower GDD than the control (P₀). In cv. Samos, only treatment P₉₀ induced earliness (Table 2). At R6 stage, cv. Flip remained the earliest while in cv. Thessaly, all the P rates scored the same GDD. On the contrary, in cvs. Samos and Ikaria, P additions caused delayed maturity (Table 2). The stages and the duration required to mature a plant are determined by many factors such as climate, genetic material, agronomic practices and soil fertility (Miller et al. 2001). It is mentioned that P fertilization is more beneficial in legumes, as it regulates the N/P ratio in the soil resulting in vigorous growth and higher yields (Li et al. 2011). According to Rasheed et al. (2010), despite the fact that P application in lentils induces earliness in flowering, it can delay maturity due to increased nitrogenase activity of active nodules and due to modification of the ratio of P to other nutrients. For the P rate × cultivar interaction, significant negative correlations between GDD and SW were found at R1, R2, R4 and R6 stages, with that for R1 stage being the strongest (Figure 2).

For the P rate × growth season interaction, all reproductive stages but R8 showed significant differences in GDD (Table 3). In general, for R1, R2 and R4 stages, P additions caused earliness, especially in 2014. However, this trend was adverse in R6 growth stage. For the above interaction, early maturity affected SW positively as negative correlations between GDD and SW were found at R2, R4 and R6 stages with that at R4 being the strongest (Figure 3).

Full maturity (R8) was influenced only by cultivars and the cultivar × growth season interaction, with cv. Flip recording the lowest GDD (1649.33 °C) and showing no significant differences for the two growth seasons (Table 4). On the contrary, cv. Ikaria showed the highest GDD, summed up to 1790.02 °C, a score very close to the upper limit (1806 °C) of GDD necessary for lentil maturity in Montana (Miller et al. 2001), an environment contrary to the Mediterranean of the present study. Early flowering and maturity genotypes have been proposed to avoid drought but because of their inability to react when there is available soil moisture they show low yields (Materne & Siddique 2009). However, no significant correlation was evident between GDD and SW for the cultivar × growth season interaction at R8 stage. It has been reported that P fertilization leads to earliness in many crops, however it was not always correlated with higher yields (Borges & Mallarino 2000).

In general, P supply stimulates root growth by increasing root biomass and length especially in P-deficient soils (Li et al. 2011). As a result, P addition has been reported to improve drought tolerance by increasing WUE in leguminous crops (Jin et al. 2014). Carbon isotope discrimination (Δ) has been extensively used as an indirect, long-term assessment of WUE and it was found to correlate with yield in C₃ species (Turner 1996; Lopes & Reynolds 2010). However, in the present study, neither cultivars nor P rates had a significant effect on Δ (Table 5) indicating that any yield differentiation among the cultivars or P treatments was not ascribed to a variation in water economy of the lentils. Previously, significant variation for Δ was found in lentil genotypes but no correlation with yield was reported (Matus et al. 1996), being in agreement with findings in other grain legumes like dry beans (Tsialtas et al. 2011).

Phosphorus effects on earliness were cultivar specific, inducing earliness in cvs. Flip and Thessaly and causing delayed maturity in Samos and Ikaria. The critical stage was found to be the beginning of flowering and earliness seems to be a factor contributing to higher yields. Neither cultivars nor P rates had a significant effect on water economy of the lentils. A proper combination of cultivar and P rate can be a means to increase lentil yields under changing Mediterranean conditions.

The authors are grateful to Dr D. Vlachostergios, HAO-DEMETER, Institute of Industrial and Fodder Plants, Larissa, Greece for providing the lentil seeds for the experimentation. We also thank Dr B. Harlow, Washington State University, School of Biological Sciences, Pullman, WA, USA for the carbon isotope determinations.