Potato is the third most important edible crop in the world (FAO 2016), and the assurance of its water requirements should be determined in order to manage its high sensitivity to water deficit (Ahmadi et al. 2010). Inappropriately adjusted irrigation and random precipitation events expose this crop to water restriction conditions affecting yield (Monneveux et al. 2013). Different models predicting potato water demand for irrigation scheduling have been reported (Raymundo et al. 2014). Most of them are based on the Penman-Monteith equation, which assumes the occurrence of stomatal closure with the concomitant absence of transpiration during dark conditions (Monteith and Unsworth 2001). It is worth highlighting that predawn water potential has been taken as a potato water status descriptor assuming that during the night, plant and soil are in equilibrium due to transpiration preclusion caused by stomatal closure (e.g. Tourneux et al. 2003; Rolando et al. 2015). Notwithstanding, the growing evidence that crops transpire important water amounts at night (Blom-Zandstra et al. 1995; Caird et al. 2007a; Schoppach et al. 2014; Resco de Dios et al. 2015) contradicts the water use modeling assumption of absent nocturnal water losses via transpiration (TRnight). One important issue that has not been fully addressed is the likely occurrence of TRnight during water restriction periods. Since water conservation through stomatal closure - even during low intensity droughts - has been documented in potato (Ramírez et al. 2016) and considering that TRnight occurrence requires an additional energy investment for plants (Caird et al. 2007b), we hypothesized a significant reduction of this process during water restriction. This preliminary work reports TRnight and nocturnal water stomatal conductance (SCnight, positive related to stomatal openness) at maximum physiological performance growing period in two advanced drought resistant potato clones (UNICA and Sarnav) under water restriction and full irrigation. Our goal was to ascertain soil water availability effects on SCnight and TRnight and their relationship with tuber yield.

Materials and Methods

Study Site, Plant Material and Management

The experiment was carried out between June 5th and September 9th 2014 (usual growing season in the region) at the International Potato Center (CIP) - La Molina experimental station in Lima-Peru (12.1° S, 77.0° W, 244 m.a.s.l.). Forty-eight tubers of UNICA (CIP code N°392,797.22) and Sarnav (CIP code N°397,077.16) cultivars were sown in pots (5.8 L) filled with a PRO-MIX and sand substratum (2:1 ratio). PRO-MIX (Premier Tech Horticulture, Canada) is a humic soil substrate with a high water retention capacity (a field capacity of 30.8% of volumetric water content). The pots were located in a glasshouse with automatized control of temperature (see specifications in Rolando et al. 2015). Each pot was fertilized with 4.0 g, 1.6 ml and 8.45 g of NH4.NO3 (31% N), H3PO4 (53% P2O5) and KNO3 (13% N and 46% K2O) respectively, distributed in 14 weekly applications. The local average daily temperature, relative humidity, solar radiation and vapor pressure deficit during the growing period were 15.28 ± 0.09 °C, 89.83 ± 0.36%, 8.03 ± 0.40 MJ m2 day−1 and 0.18 ± 0.0072 kPa respectively (HOBO U12 Outdoor/Industrial Data Logger, Onset Computer Corporation, Bourne, MA, USA).

Long-term water restriction was induced after tuber initiation; i.e. when stolon tips were twice the diameter of subtending stolon, determined through sequential harvesting of eight plants per cultivar. Tuber initiation occurred at 39 and 34 days after planting for UNICA and Sarnav. Following a completely randomized design in a glasshouse, 16 plants per cultivar were randomly assigned to control (field capacity) and water restriction treatments, 8 plants per treatment. Plants submitted to water restriction were irrigated to 50% of field capacity, guided by gravimetric assessment of all the plants, as described by Rolando et al. (2015). The irrigation was carried out twice per week up to harvest.

Physiological Measurements

Physiological measurements were taken at 70 days after planting (DAP, i.e. 31 and 36 days after tuber initiation onset for UNICA and Sarnav respectively), time at which both cultivars reached their maximum canopy cover, tuber bulking rate and photosynthetic performance (Rolando et al. 2015). Nocturnal water stomatal conductance (SCnight) was assessed both at early-night (8:30 pm – 10:30 pm) as well as at predawn (2:30 am - 4:30 am) using a portable photosynthesis system (LI6400 TX model, LI-COR, Lincoln, NE, USA). In both periods the following parameters were fixed in the cuvette: boundary layer = 9.3 mol m−2 s−1, CO2 concentration = 400 ppm and air temperature = 20 °C. Nocturnal transpiration (TRnight) was assessed gravimetrically in tandem with SCnight, by covering the soil surface of the pots with a transparent plastic to preclude evaporation losses. TRnight was estimated as the pot weight difference from 6:30 pm to 6:00 am of the following day. Total day transpiration was assessed through the weight difference in a period of 24 h. Tuber fresh biomass was measured 96 DAP, weighting all the tubers per plant.

Statistical Analysis

The effect of cultivars and watering treatments on SCnight and transpirations (nocturnal and daily) were assessed through an ANOVA, followed by a Fischer’s Least Significant Difference (LSD) test. A paired two sample t-Test was used to compare periods of SCnight (early-night and predawn). Linear regression analyses were performed – as described by Resco de Dios et al. (2016) - to explain tuber fresh biomass with SCnight and TRnight taken at maximum physiological performance. All the statistical analyses were run using SAS 191 v.8.02 software (SAS Institute, Cary, NC, USA).

Results and Discussion

Nocturnal Stomatal Conductance and Transpiration

Significant differences for SCnight at early-night and predawn period were observed between cultivars and water treatments, respectively (Table 1). The interaction cultivars × water treatment was significantly different only at predawn period (Table 1). The paired two sample t-Test between early-night and predawn SCnight showed differences (at P<0.01) only for UNICA under well-watered conditions (Fig. 1a). The highest value of SCnight was detected at predawn under well-watered conditions in UNICA (see Fig. 1a). This value was in the range of predawn SCnight values reported for other Solanaceae species (0.04–0.10 mol H2O m−2 s−1; Caird et al. 2007a). It has been suggested that predawn SCnight is related to the degree of circadian control which determines plant readiness to anticipate the beginning of the day (Resco de Dios et al. 2016) with important influences in the control of TRnight (Resco de Dios et al. 2015). This readiness is, in turn, associated to the circadian resonance, which reflects the degree of connection of endogenous rhythms with environmental light-dark cycles (Dodd et al. 2005) determining the success of a specie by fitness advantage (McClung 2006).

Table 1 Results of two-way Analyses of Variance testing the effect of cultivars, water treatments and their interaction in nocturnal transpiration (TRnight), total daily transpiration (TRdaily), early-night and predawn nocturnal stomatal conductance (SCnight). *P < 0.05, **P < 0.01, n.s. P > 0.05
Fig. 1
figure 1

Average nocturnal stomatal conductance (a) and transpiration (b) under two water status in UNICA and Sarnav cultivars. Different capital letters in A and B mean significant differences among early-night stomatal conductance and total daily transpiration using LSD test. Different lower case letters in A and B mean significant differences among predawn stomatal conductance and nocturnal transpiration using LSD test

Significant differences for TRnight and total daily transpiration were observed between cultivars and watering treatments, respectively (Table 1). The interaction cultivars × watering treatment was significantly different only in TRnight (Table 1). The average percentage of TRnight in relation to the total daily transpiration under non-restricted and water restricted conditions were: 10.6 ± 0.5% and 4.2 ± 4.7% for UNICA, and 7.7 ± 0.7% and 6.4 ± 1.5% for Sarnav, respectively (Fig. 1b). The observed values under non-restricted watering conditions were in the range of other Solenaceae crops (7.7–10.6%; Caird et al. 2007a), but they were lower than the values reported for roses (35–55%, Blom-Zandstra et al. 1995) and wheat genotypes (13.8–54.9%, Schoppach et al. 2014) assessed under no water restriction. Our night-time vapor pressure deficit (VPDnight) during the TRnight assessment ranged from 0.10 to 0.23 kPa which are low values, typical of the Peruvian Coast characterized by high atmospheric humidity. Because VPDnight is considered one of the main drivers of TRnight (Schoppach et al. 2014), we hypothesize that potato could potentially show higher TRnight in regions with higher VPDnight and genotypes with larger predawn SCnight values (see Fig. 1a). Given the occurrence of TRnight under water restriction (an average of 5.4 ± 2.6% of the total daily transpiration, see Fig. 1b) and the additional energy investment it entails (Caird et al. 2007b), TRnight in potato could be functionally involved in other physiological processes. In other species, TRnight has been linked to mass flow enhancing soil nutrients uptake (Caird et al. 2007b; Matimati et al. 2014), carbohydrates exportation associated to dark respiration (Marks and Lechowicz 2007), and other processes (Schoppach et al. 2014) that require further research in potato.

Fig. 2
figure 2

Linear functions fitted to explain tuber fresh biomass (TFB, g plant−1) and nocturnal transpiration (TRnigth, g H2O plant−1 night−1) variability through their relationship with nocturnal stomatal conductance (SCnight, mol H2O m−2 s−1) measured at early-night (open circles, black continuous line) and predawn (black circles, grey dashed line). *P < 0.01, **P < 0.001, ***P < 0.0001

Relationship between Tuber Yield and Physiological Variables

The variables that better explained fresh tuber biomass were TRnight and predawn SCnight (Fig. 2a,b), whereas TRnight was mostly related to predawn SCnight (Fig. 2c). Our results agree with recent findings (Resco de Dios et al. 2016) which support the novel hypothesis that circadian resonance play a crucial role in maximum assimilation, biomass accumulation and fitness. The high C allocation to tubers in the potato crop (harvest index ranged between 0.67–0.87, Condori et al. 2010), and the capacity to anticipate daylight leading to a high predawn SCnight, might hint a fast-adapting to starting and quenching photosynthesis process in improved genotypes.


The high TRnight in potato, even in conditions of low VPDnight and soil water deficit, is not consistent with the occurrence of stomatal closure at night, an assumption that should be re-analyzed for water balance modeling and water status estimations in this crop. The preliminary results reported in this study emphasize the necessity to understand the physiological processes in which TRnight is involved and why potato plants transpire at night even under water restriction conditions. Future studies must assess SCnight and TRnight throughout the growing season (not restricted to the season of maximum physiological performance), with more genotypes, different water restriction levels and environments. The hypothesis that circadian resonance, through daylight anticipation, might be involved in regulating tuber biomass accumulation in this crop is also a pending research issue to ascertain the usefulness of predawn SCnight in breeding programs under well water conditions, as a trait related to tuber yield.