Plant material and experimental conditions
This study was conducted at the Centro de Estudios Avanzados en Fruticultura (CEAF), Rengo, Chile (34°19’S 70°50’W). The climate corresponds to a warm temperate climate with an annual precipitation of 350 mm, an average maximum temperature in January of 28.5°C, and an average minimum temperature in June of 2°C. The reference evapotranspiration for the summer period (December to February) is 465 mm (CIREN 2023).
Fruit trees included in the trial were genetically distant, have a wide range of drought tolerance, and are economically important in different edaphoclimatic locations. One-year-old plants of avocado (Persea americana Mill.), mandarin (Citrus reticulata Blanco), pomegranate, fig, Rootpac®40 rootstock (Prunus dulcis [Mill.] DA Webb x P. persica Batsch), and Rootpac®20 rootstock (Prunus besseyi Bailey × P. cerasifera Ehrh) were obtained from a commercial nursery (Table 1). We worked with grafted mandarin and avocado plants to represent the way these crops are traditionally grown.
Table 1
List of the studied cultivars and rootstocks, tree species origin, and climate of the center of origin.
Scion | Rootstock | Specie or genera origin and climate | Reference |
Pomegranate cv. 'Wonderfull' (Punica granatum L.) | Own-rooted | Transcaucasia and Central Asia – Subtropical to desertic | (Chandra et al., 2010) |
Fig cv. 'Black Mission' (Ficus carica L.) | Own-rooted | Middle East – Arid to mild-temperate | (Mars, 2003) |
| | Central and southwest Asia – Desertic to subtropical (almond or peach respectively) | |
Rootpac®40 (Prunus dulcis [Mill.] DA Webb x P. persica Batsch) | Own-rooted | (Kester et al., 1991, Ladizinsky, 1999) |
Rootpac®20 (Prunus besseyi Bailey × P. cerasifera Ehrh) | Own-rooted | From the Balkans to the Caucasian mountains in southwest Asia – Temperate to semiarid (P. cerasifera) | (Horvath et al., 2008) |
Mandarin cv. 'Orogrande' (Citrus reticulata Blanco) | ‘Carrizo’ (Citrus sinensis L. Osb.× Poncirus trifoliata L. Raf.) | Southeastern Asia and Australia – Subtropical (Citrus spp) | (Scora, 1975) |
| | Mesoamerica – Tropical to subtropical | |
Avocado cv. 'Hass' (Persea americana Mill.) | ‘Mexícola’ (Persea americana Mill.) | (Galindo-Tovar et al., 2008) |
Plants were transferred to 20 L pots filled with a mixture of 1:1 peat/perlite supplemented with Basacote Plus 9M at a 6 g L− 1 as a controlled-release fertilizer (BASF, Limburgerhof, Germany). After the transplant, plants were pruned to activate multiple growth points. Plants were grown for 30 days in a shade house (50% sunlight) and were then transferred to field conditions (full sunlight). An acclimation period of two weeks was allowed before the experiment started. The irrigation system consisted of two drippers per plant with a flow rate of 2 L h− 1. Pots were covered with black plastic bags to avoid evaporation. Sixteen uniform and healthy plants of each species were selected for all evaluations. Of the sixteen plants, four plants were used to determine the initial biomass, and the remaining 12 plants were used for final biomass and physiological measurements during the water deficit period.
Irrigation treatments
Plants were subjected to two irrigation conditions: well-watered (WW) and water deficit (WD), in a completely randomized experimental design. Each pot was saturated with tap water, allowed to drain, and covered with plastic bags to avoid evaporation for 24 h. After, a similar initial weight of each container was reached (ca. 11 kg) and recorded. This was established as 100% of soil water content (SWC) and considered as field capacity. WW plants were irrigated three times per week adding tap water until 100% of the SWC for each pot. For WD the irrigation was withheld for 45 days, weighing each container three times per week to determine the water consumption of every single plant. During this period all the containers were filled with the volume of water necessary to reach the same water content of the plant with the lowest transpiration rate which, in turn, was not irrigated. Therefore, after the irrigation, all the WD containers had the same SWC (Opazo et al., 2019, Opazo et al., 2020) and the progression of the water deficit was similar between species. The minimum water content in the pot (0% SWC) was determined by drying the substrate in an oven at 100°C until constant weight. The SWC was calculated for each container as Eq. 1 (Supp. Figure 1):
$$\text{S}\text{W}\text{C} \left(\text{%}\right)= \frac{\text{D}\text{S}\text{W}-\text{D}\text{W}}{\text{M}\text{S}\text{W}- \text{D}\text{W}}; [\text{E}\text{q}.1]$$
where DSW is the daily substrate weight, DW is the substrate dry weight reached in the oven and MSW is the maximum substrate weight at field capacity.
Relative transpiration (RT) rate and residual soil water content (RSWC)
The point of reduction in transpiration under WD was calculated by the relationship between the relative transpiration rate (RT) and the fraction of transpirable substrate water (FTSW) (Sinclair and Ludlow, 1986). The transpiration rate corresponds to the daily amount of transpired water under WD, divided by the average daily transpiration of the WW plants for each species. The fraction of transpirable substrate water corresponds to the fraction of water inside the container that plants can use to transpire. The RT of each plant was divided by the respective mean RT of each plant during the period when the soil was still well-watered to normalize the initial values (Sinclair and Ludlow, 1986). According to Bindi et al. (2005), the initial point for stress (FTSWthreshold) is around a RT value of 0.9. Then, RT was adjusted to a logistic equation as in Eq. 2:
$$\text{R}\text{T}= \frac{1}{1+{\alpha }\text{*}{e}^{-{\beta }\text{*}\text{F}\text{T}\text{S}\text{W}}}; [\text{E}\text{q}.2]$$
where α and β are constants to be determined for each plant related to the curvature of logistic regression.
At the end of the experiment, containers were weighed before the plant harvest. After, the fresh weight of leaves, stems, and roots, plus the weight of an empty container, were discounted to the initial weight to obtain the substrate weight at this moment on every WD plant. That substrate weight was compared with the dry weight (70°C) to determine the residual water content in WD plants.
Stomatal conductance, net photosynthesis, stem water potential, and osmotic potential
Stomatal conductance (gs) and net photosynthesis (A) were measured weekly on fully active and expanded leaves between 9:30 and 11:00 a.m. on the day before irrigation. One leaf per experimental unit was measured using a portable photosynthesis system (model CIRAS-2, PPSystem, Hitchin, UK) equipped with a 2.5 cm2 LED lighting cuvette (model CIRAS PLC, PPSystem). Measurements were made at 25°C, with photosynthetic active radiation of 1000 µmol PAR m− 2 s− 1, CO2 concentration of 400 µmol mol− 1, and relative humidity of 50%. Midday stem water potential (Ψstem) was measured on similar leaves and the same-day gas exchange measurements. One leaf per experimental unit was covered with plastic bags coated with aluminum foil to stop transpiration and allow them to balance with the stem water potential for at least 2 hours before measuring. The measurement was made at solar noon (between 13:00 and 15:00 local time) with a Schölander pressure chamber (Schölander et al., 1965).
Relative stomatal conductance (gsrel) and relative net photosynthesis (Arel) were estimated by the ratio between measurements made on WD plants and average WW. Negative values of A were assumed equal to 0. These relative traits were related to FTSW in the same manner as described in Eq. 2 to estimate gas-exchange thresholds. With the same purpose a relative stem water potential (Ψstem rel) was calculated as in Eq. 3:
$${{\Psi }}_{\text{s}\text{t}\text{e}\text{m} \text{r}\text{e}\text{l}}= \frac{\text{W}\text{D} {{\Psi }}_{\text{s}\text{t}\text{e}\text{m}}-\text{m}\text{i}\text{n}\text{i}\text{m}\text{u}\text{m} {{\Psi }}_{\text{s}\text{t}\text{e}\text{m}}}{\text{a}\text{v}\text{e}\text{r}\text{a}\text{g}\text{e} \text{W}\text{W} {{\Psi }}_{\text{s}\text{t}\text{e}\text{m}}- \text{m}\text{i}\text{n}\text{i}\text{m}\text{u}\text{m} {{\Psi }}_{\text{s}\text{t}\text{e}\text{m}}}; [\text{E}\text{q}.3]$$
where minimum Ψstem corresponds to the more negative Ψstem value detected during the experimental period in WD plants for each species.
Osmotic water potential (Ψo) was measured after 45 days of treatment. Fully active and expanded leaves were collected and kept in distilled water for 24 h and then frozen at -80°C until measurement. Leaves were thawed and put in a syringe to press them and extract their sap. The osmolality was evaluated with an osmometer (Osmomat 3000, Gonotec GmbH, Germany). The osmotic potential was obtained using the Van’t hoff equation, where osmotic potential = C*T*R, where C is the osmolality (mOs mol kg− 1 H2O), T the absolute temperature and R is the gas constant (0.00831 kg MPa mol− 1 K− 1).
Whole plant hydraulic conductance
Whole plant hydraulic conductance (Kpl; Kg [H2O] MPa− 1 m− 2 s− 1) was determined at day 38 of the water deficit period (plants severely stressed) on four plants per treatment. The whole plant transpiration (E; Kg [H2O] m− 2 s− 1) were calculated from Eq. 4:
$$\text{E}= \frac{{\Delta }\text{W}}{\text{L}\text{A} \text{*} {\Delta }\text{t}}; [\text{E}\text{q}.4]$$
where Δt (seconds) was the interval time between dawn and midday, ΔW was the weight container change (kg) between pre-dawn and midday, and LA was leaf area (m2), which was determined at the end of the experiment (7 days later). Pre-dawn and midday leaf water potential were measured with a pressure chamber. Kpl was estimated as in Eq. 5 (Tsuda and Tyree, 2000):
$${\text{K}}_{\text{p}\text{l}}= \frac{\text{E}}{{{\Psi }}_{\text{s}\text{o}\text{i}\text{l}}-{{\Psi }}_{\text{l}\text{e}\text{a}\text{f}}}; [\text{E}\text{q}.5]$$
Root hydraulic conductivity
The root hydraulic conductivity (Lp) was determined by a High-Pressure Flow Meter (HPFM, Dynamax, Houston, TX, USA) at the end of the experiment, according to Tyree et al. (1995) on six plants per treatment. The night before measurements, WD plants were fully irrigated to recover the water columns, reducing artifacts due to cavitation (Alsina et al., 2011). At the end of the water deficit period, Lp measurements were carried out in the whole root system below the rootstock/scion junction in grafted plants, and at 10 cm over the substrate’s surface in non-grafted plants, twice per plant. Subsequently, this value was normalized by the dry weight of the whole root system (Vandeleur et al., 2014).
Iso-anisohydric behavior
On days 24, 31, and 38 of water deficit, leaf water potential was measured at pre-dawn (Ψpd) and midday (Ψmd; between 13:00 and 15:00 h) on four plants per treatment. A modified hydroscape area (Meinzer et al., 2016) was estimated by the method proposed by Johnson et al. (2018). In this method, the hydroscape area (HA: MPa2) is calculated as the area of a polygon formed by the 1:1 lines (Ψmd = Ψpd) and the leaf midday water potential. A greater HA will be associated with a more anisohydric stomatal regulation on leaf water potential. The polygon areas were obtained through digital image analysis with the ImageJ software (Schindelin et al., 2015).
Isotopic composition of 13C
To determine the leaf isotopic composition of 13C (δ13C; ‰) fully expanded and sun-exposed leaves were sampled on day 45 of water deficit. Samples were dried at 60°C in a forced-air oven until constant weight and then ground in a mill to a homogeneous fine powder. Two subsamples from each sample were weighed with an analytical balance (Precisa 125A) in tin capsules for δ13C measurements. Each leaf sample isotopic composition was determined using standard procedures at the Stable Isotope Laboratory at the Faculty of Agronomic Sciences (University of Chile), with an INTEGRA2 isotopic ratio mass spectrometer (IRMS) (Sercon Ltd. Cheshire, UK), with a precision of 0.3‰ for δ13C.
Wood density and stomatal density
Wood density and stomatal density were measured after 45 days of treatment. Wood density was estimated as the ratio between the dry mass of a trunk segment and its maximum fresh volume. Segments were placed in distilled water for 24 hours, then the bark was removed, and their volume was determined through a dimensional method by measuring their length and diameter (average between the beginning, middle, and end of the segment). Subsequently, the segments were dried in a forced-air oven at 80°C until constant weight for dry mass measurement. Stomatal imprints were made by applying a nail varnish on the abaxial surface of the leaves, avoiding the midrib and the leaf margin. After drying, the nail varnish film was gently peeled off using transparent adhesive tape and was fixed on a clean labeled microscope slide (Kardel et al., 2010). The stomatal imprints were analyzed with a light microscope (Olympus BX43, Olympus, Hamburg, Germany). For each imprint, two images were taken in different zones. Stomatal density (number of stomata per mm2) was assessed by counting all the stomata of the image (known area) and extrapolating to 1 mm2.
Whole-plant water-use efficiency and growth
At the beginning and end of the experiment, plants were harvested and divided into leaves, stems, and roots. The dry weight of each plant part was determined after placing the samples in an oven at 70°C until to reach a constant weight. Growth for each plant part was calculated as the means of the difference between the final biomass of leaves, stems, and roots and the average of the initial biomass of each species. The specific leaf area (cm2 g− 1) was measured by scanning leaves and then analyzed with ImageJ software (version 1.51j8 NIH) with a reference area (O’Neal et al., 2002) and then dried in an oven at 70°C until reaching a constant weight. The total leaf area per plant was estimated by multiplying the specific leaf area and the total dry weight of the leaves.
Whole-plant water-use efficiency (WUEwp) was calculated as the difference between the total biomass at the end of the water deficit period minus the average biomass at the beginning of the experiment, divided by water consumption of the respective period, as in Eq. 6.
$${\text{W}\text{U}\text{E}}_{\text{w}\text{p}}= \frac{\text{F}\text{i}\text{n}\text{a}\text{l} \text{d}\text{r}\text{y} \text{b}\text{i}\text{o}\text{m}\text{a}\text{s}\text{s}-\text{I}\text{n}\text{i}\text{t}\text{i}\text{a}\text{l} \text{d}\text{r}\text{y} \text{b}\text{i}\text{o}\text{m}\text{a}\text{s}\text{s}}{\text{T}\text{o}\text{t}\text{a}\text{l} \text{w}\text{a}\text{t}\text{e}\text{r} \text{c}\text{o}\text{n}\text{s}\text{u}\text{m}\text{p}\text{t}\text{i}\text{o}\text{n}}; [\text{E}\text{q}.6]$$
Statistical analysis
Differences in FTSWthreshold, gas exchange threshold, SWC to FTSW = 0, Lp, Kpl, biomass, WUEwp, and δ13C were tested using irrigation, species, and their interactions as fixed factors. Heteroscedastic models were used when necessary, followed by a Ficher-LSD post hoc analysis when appropriate. For analysis, a significance level of 0.05 was set. A principal component analysis (represented by a biplot) was carried out as an exploratory analysis of the associations between traits, and the similarity between species under well-watered and water deficit conditions. All the statistical analyses were made using InfoStat (version 2016e, Universidad de Córdoba, Córdoba, Argentina) statistical software (Di Rienzo et al., 2011)d v 4.1.1. (R-Core-Team, 2020).