Mulching improved soil fertility, plant growth and productivity, and postharvest deficit irrigation reduced water use in sweet cherry orchards in a semi-arid region

ABSTRACT In the Okanagan Valley, sweet cherry production has expanded to higher latitudes due to climate change, but the availability of irrigation water is limited in this semi-arid region. Postharvest deficit irrigation (PDI) and organic mulches can reduce water use in orchards, but their interactive effects on soil fertility, water relations, and crop performance in new orchard environments are unknown. In a randomized block split-plot design, full irrigation (100%) or PDI (72–76% of full irrigation) was applied to the main plots, and mulches (compost, woodchips, bare) were subplots at three sites. Compost increased soil organic matter, nutrients, pH, and electrical conductivity over three seasons at all sites. Woodchips increased tree growth and foliar P and Mn, while compost increased some fruit quality attributes, and foliar P compared to bare soil. Relative to full, PDI saved 24–28% irrigation water after harvest per season at each site without affecting soil moisture and chemical properties, stem water potential, or crop performance, or interacting with mulch effects. These results suggest that in this semi-arid cherry growing region mulches are a promising strategy to maintain soil moisture and improve soil fertility and crop performance, and PDI can reduce water use after harvest without affecting commercial production.


Introduction
The Okanagan Valley of British Columbia, Canada, which spans latitudes 49.00° to 50.21° N, is an important fruit-producing region. In past decades, the desert climate of the south Okanagan allowed more intensive tree fruit and vine production than the cooler, but semi-arid climate of the north Okanagan, because crops such as sweet cherry are sensitive to late spring frosts and winter temperatures below −30°C . However, recent climate models indicate that cherry trees can be grown in more northern and higher elevation areas of the Okanagan Valley, due to warmer temperatures and an increasing number of frost-free days brought by climate change . Growers in these areas are converting native grasslands and previous grazing areas to sweet cherry production. Low levels of precipitation in the Okanagan Valley during the growing season make irrigation essential to the commercial cultivation of sweet cherries. However, the availability of an adequate and regular water supply is a major concern that is compounded by competition for freshwater between agricultural production and a rapidly growing urban population in the watershed (Neilsen et al. 2006. Thus, the development of orchard management practices that increase the water use efficiency of cherry orchards, such as mulching or deficit irrigation, would help ensure that the expansion of cherry production in the Okanagan Valley is sustainable. Because sweet cherry production typically occurs in semi-arid regions globally, the development of such water-conserving practices would be relevant to orchard production in other parts of the world. In addition to reducing evaporation from the soil surface, organic mulches increase soil organic matter contents, water holding capacities and nutrient availabilities (Forge et al. , 2008(Forge et al. , 2013Hannam et al. 2016), which in turn benefit the functioning of root systems and tree water relations Forge et al. 2008;Sas-paszt et al. 2014). Although numerous studies have demonstrated that these benefits of mulches to soil culminate in improved apple tree growth or fruit production TerAvest et al. 2011;Forge et al. 2013;Jones et al. 2020), few such studies have been conducted with sweet cherry (Watson et al. 2017), and mulching with organic materials has not been a common practice in sweet cherry orchards in the Okanagan Valley. Postharvest deficit irrigation (PDI) is a water-saving technique where irrigation is applied during drought-sensitive phenological stages of a crop, and reduced during drought-tolerant stages, such as the postharvest period of perennial fruit trees (Ruiz-Sanchez et al. 2010). Postharvest deficit irrigation has been shown to address the issue of limited water supply and to increase water use efficiency, without negative effects on fruit quality and yield (reviewed by Ruiz-Sanchez et al. 2010;Behboudian et al. 2011;Chai et al. 2016). Research on sweet cherry is limited, but a few studies in Spain using PDI (50% reduction) showed sweet cherry fruit yield was maintained at a similar level to fully irrigated trees (Marsal et al. 2010;Behboudian et al. 2011). While organic mulches can influence soil water relations and elements of soil fertility that, in turn, affect tree water relations, the interactive effects of PDI and mulching have never been examined. The objectives of this study were to extend current knowledge of the effects of mulches on soil physicochemical properties, tree growth and fruit production, and to determine if annual mulching moderates the effects of PDI on tree performance in sweet cherry orchards. The study examined the influence of two irrigation treatments (full irrigation and PDI [72-76% of full]) and soil mulching treatments (compost, woodchips, and bare) on selected soil chemical properties, as well as the water relations, growth and productivity of cherry trees over three growing seasons (2015, 2016, and 2017) in three sweet cherry orchards in the Okanagan Valley. We hypothesized that mulched soil would have greater soil moisture and organic matter content compared to bare soils. We hypothesized that stem water potential, tree growth, and fruit yield and quality would be similar between sweet cherry trees growing under 72-76% PDI and full irrigation treatments.

Description of the experimental sites
The study sites were located at three newly-planted orchards in the Okanagan Valley of British Columbia: two orchards were at sites never before used to grow tree fruits (Sites 1 and 2) and one was at a site with a long history of apple production (Site 3). Site 1 (50° 14' N/ 119° 8' W) was a former pasture land that had been converted into an orchard in April 2015 when sweet cherry trees ['Stacatto' (Prunus avium L.) on Mazzard [(P. avium) rootstock] were planted. Site 2 (50° 14' N/ 119° 7' W) was a former dairy farm that was planted with sweet cherry trees ['Skeena' (P. avium) on Giesela 6 (P. cerasus × P. canescens) rootstock] in spring 2014. Unlike the new orchards, Site 3 (49° 51' N/ 119° 23' W) had been used for the production of apples before it was planted with sweet cherry trees ['Sentennial' (P. avium) on Mazzard (P. avium) rootstock] in spring 2013. The first year of cherry harvest for Sites 2 and 3 was 2016, while at Site 1 the first year of harvest was 2017. The soil type at Sites 1 and 2 was sandy loam and loamy sand at Site 3. The soil series at all three sites was calcic haploxerols. Experiments were established in June 2015 and maintained at all three sites for the 2015, 2016, and 2017 growing seasons. In each growing season, trees at all three sites were fertigated with 285 g N tree −1 as urea at the end of March; 272 g tree −1 of 20-20-20 (N-P-K) blend in mid-May; and 163 g N tree −1 as calcium nitrate three times from mid-May to the end of June; and 0.95 g Mg tree −1 , as magnesium sulphate, in May. Standard nutrition and pesticide regimes were practiced at all sites in each growing season (BC Tree Fruit Production Guide 2021).

Experimental design and treatments
The experiment at each site was a randomized block split-plot design with three subplots contained in each of twelve whole plots. Each subplot had two measurement trees flanked by two guard trees, and was randomly allocated one of the three mulch treatments. Distance between trees at Sites 1, 2, and 3 were 2.2 m, 2.4 m, and 2.7 m, respectively. Lengths of treatment rows at Sites 1, 2, and 3 were approximately 52.8 m, 57.6, and 64.8 m, respectively. The twelve whole plots were grouped into six blocks; within each block, the two irrigation treatments were randomly allocated to one of the two whole plots, resulting in six replicate blocks of each irrigation whole-plot treatment and a total of twelve subplots of each mulch treatment. The two irrigation treatments were full irrigation (100%) and postharvest deficit irrigation (PDI). Postharvest deficit irrigation involved a 26%, 24%, and 28% reduction of full irrigation at Site 1, 2, and 3, respectively. At Sites 1 and 2, full irrigation was applied using driplines with a drip spacing of 0.46 m between emitters at a rate of 1 L per hour (LPH), while the PDI treatment was applied using driplines with 0.61 m drip spacing between emitters at the same rate, thus reducing applied water by 26% at Site 1 and 24% at Site 2. At Sites 1 and 2, trees were irrigated daily for six hours during the growing season, and 4 h wk −1 after harvest. At Site 3, 72 LPH microsprinklers were used for full irrigation, while 52 LPH microsprinklers were installed for PDI to reduce applied water by 28%. Here, trees were irrigated a total of 8 to 9 h wk −1 during the growing season and 4 to 5 h wk −1 after harvest. Placement of irrigation lines and emitters in both irrigation treatments in relation to the tree was recommended by local production guidelines (BC Tree Fruit Production Guide 2021). Full irrigation was used in all plots from early spring until harvest; then PDI was implemented right after harvest until the end of the growing season (first week of October) in half of the whole plots. Because the first harvest took place in different years depending on the site, PDI was implemented on 1 August 2016 at Sites 2 and 3, and again on 8 August 2017 at Site 2 and 18 August 2017 at Site 3. At Site 1, PDI was first implemented on 24 August 2017. None of the sites received PDI in 2015. The three mulch treatments applied to subplots were: 1) compost mulch, 2) wood chip mulch, and 3) non-mulched (bare) control. The compost (GlenGrow) was made from yard waste by the City of Kelowna, BC, Canada. Douglas-fir woodchip mulch was sourced from local landscaping companies in Vernon, BC, to apply at Site 1 and 2, and Kelowna, BC to apply at Site 3. The mulches were first applied in June of 2015 to the surface of the soil in a 1.5 m-wide band centered on the tree row to a depth of 0.05 m, and re-applied at the same rate in May of 2016 and 2017. The nutrient composition of the compost was supplied by the City of Kelowna, while nutrient analyses for the Douglas-fir woodchip mulches were done by the BC Ministry of Environment Analytical Lab (Victoria, BC) (Table S.1). In each year of application, the total input per subplot was estimated to be 106.9 Mg ha −1 C and 7.25 Mg ha −1 N from the compost at all three sites, 239.6 Mg ha −1 C and 2.1 Mg ha −1 N from woodchip mulch at Sites 1 and 2, and 235.3 Mg ha −1 C and 1.7 Mg ha −1 N from woodchip mulch at Site 3 (Table S.1). Based on previous studies and guidelines predicting N mineralization as a function of C: N ratios of composts incorporated into soil (Gale et al. 2006), we assumed that the rate of N mineralization in the compost would have been in the range of zero to 10% of total nitrogen. Considering that the compost was not incorporated into the soil upon annual reapplication, it is likely that the actual rate of mineralization was at the lower end of the range of estimated N mineralization rates for the compost. Anecdotally, we visually noticed that both the compost and woodchip mulches appeared to breakdown the most after the first application in 2015, and upon reapplication in 2016 and 2017 we noticed the mulches decomposed more slowly thereafter.

Soil moisture
In November 2015 soil moisture probes (Decagon Devices, Pullman, WA) were installed in each subplot between the two measurement trees, at a depth of 30 cm, in three of the six blocks at each site (Figure S.1). Probes were connected to automated data loggers (Em50 or Em5b loggers; Decagon Devices) set to record soil moisture every 1 h; the data were manually downloaded bi-weekly. In order to assess the effect of irrigation and organic management practices on soil properties, it is common practice to take soil measurements at a 30 cm depth (Olson and Al-Kaisi 2015). Due to issues with missing soil moisture data in 2016 at all sites and in 2017 at Site 1, there were only enough data to study the interactive effect of Irrigation × Mulch at Sites 2 and 3 in August and September of 2017.

Soil sampling and analyses
Initially, soils were sampled in May 2015 to determine the baseline soil characteristics of the selected experimental sites (Table S.2). Then, soil and roots were collected in each experimental subplot in October 2015, 2016, and 2017. A total of four soil cores were taken 30 cm from each measurement tree of each subplot with a 2-cm diameter soil corer to a depth of 30 cm and combined in the field to form a composite sample representing each subplot. When sampling mulch treatments, compost and woodchips were pushed aside to expose the soil prior to taking cores. Soils were transported to the lab in coolers. They were sieved through a 5-mm sieve before use for the soil physicochemical analyses. Storage conditions and analysis for subsamples are described in the following subsections.

Soil physicochemical properties
Immediately after sieving, 1 g of fresh soil from each sample was placed into a pre-weighed tin and dried at 105°C for 48 h to determine gravimetric water content (GWC). After being air-dried and sieved again (<1 mm), the soil samples were subjected to the following in-house analyses: pH and electrical conductivity (Schofield and Taylor 1955); and permanganate oxidizable carbon (POXC, Weil et al. 2003). Dried soil was sent to A&L Laboratories (London, Ontario, Canada) in 2015 and to the British Columbia Ministry of Environment Analytical Lab (Victoria, B.C.) in 2016 and 2017 for the following soil nutrient analyses: total C and N (Thermo Scientific 'Flash 2000' combustion elemental analyzer); available nitrate and ammonium (OI Analytical 'Alpkem FSIV' segmented flow autoanalyzer); P (Agilent Cary-60 UV/visible spectrophotometer); cation exchange capacity (CEC) and exchangeable bases (K, Mg, Ca, and Na) (Dual-view ICP spectrometer; Prodigy, Teledyne/Leeman, Hudson, New Hampshire); and soil organic matter (Kalra and Maynard 1991;Kalra 1998).

Stem water potential (Ψ stem )
The midday (10 am -12 pm) Ψ stem of cherry trees was used to evaluate the effect of irrigation and mulch treatments on the water status of the trees in 2016 and 2017 for Sites 2 and 3, but only in 2017 for Site 1 (PDI was not established at this site until August 2017). Four out of six blocks at each site were randomly selected and measured once every 2 weeks. Two leaves per tree (two trees per subplot), located on the shaded side of each tree, were wrapped with black plastic and aluminum foil while attached to the tree to reduce the effect of transpiration, and left for an hour to allow equilibration with the stem before excision. After leaf excision, Ψ stem was measured using a Scholander pressure chamber (Model 3005: Soil moisture Equipment Corp., Santa Barbara, CA, USA).

Tree trunk cross-sectional area and leaf area
Using a digital caliper, annual measurements of trunk diameter were made on the two measurement trees in each subplot during November of each sampling year, and the trunk cross-sectional areas subplot −1 calculated by averaging the TCSA of the two measurement trees. To determine the average leaf area, 10 leaves were collected from each of the two measurement trees in each subplot from the mid-third portion of extension shoots of the current year's growth in mid-summer (mid-July to early August) of each sampling year. The area of each collected leaf was then measured using a leaf area meter (LI-3000, LI-COR Inc.).

Foliar nutrient content
In each sampling year, the leaf samples collected for leaf area were subsequently used to determine the nutrient content of cherry tree foliage in each subplot. The samples were oven-dried at 65°C for 24 h and then ground in a Wiley Mini-Mill (Thomas Scientific). Total N, C and S were quantified using a Thermo Scientific 'Flash 2000' combustion elemental analyzer; for Ca, P, K, Mg, Al, B, Cu, Fe, Mn and Zn, the samples (250 mg subplot −1 ) were analyzed using a dual-view ICP spectrometer (Prodigy, Teledyne/Leeman, Hudson, New Hampshire) (Kalra, 1998) following digestion in concentrated nitric acid.

Fruit yield and fruit quality analysis
Fruit was harvested and total yield was measured in 2016 and 2017 for Sites 2 and 3, but only in 2017 for Site 1. Fruit quality analysis was completed as described by Neilsen et al. (2014). In brief, the following measurements were taken: average fruit weight; percentage of fruit with split skin; fruit firmness (FirmTech II, Bioworks, Stillwater, OK); stem pull force (SPF) (Dart FGV-5X digital force gauge); fruit color (Michigan State University 1 to 5 scale chart) (Michigan State University, East Lansing, MI); soluble solids concentration (SSC) (Digital Refractometer Model PR-101; AD Scientific Instruments, Keene, NM); and titratable acidity (TA) (Model 719S Titrino, Herisau, Switzerland).

Statistical analyses
For soil and tree growth data the mulch treatments were in place all three years of the study, whereas the PDI treatment was implemented in 2016 and 2017 at Sites 2 and 3, and only in 2017 at Site 1. Therefore, the effects of Irrigation, Mulch and the Irrigation × Mulch interaction were first examined using repeated measures two-factor split-plot ANOVA with data from Sites 2 and 3 in 2016 and 2017 for all parameters except stem water potential (Ψ stem ), average monthly soil moisture content and fruit quality. Irrigation and Mulch were designated fixed effects, Site and Block were random effects, and the two years of measurement represented the repeated measures in this ANOVA model. Because neither the Irrigation nor Irrigation × Mulch interaction effects were significant for soil and tree growth parameters, a more robust repeated measures blocked one-factor ANOVA that included data from all three years was also conducted for all parameters that were measured in all three years: soil chemical variables, foliar nutrients, TCSA, and leaf area. Mulch was the main fixed factor, Site and Block were treated as random factors, and the three years of measurement represented the repeated measures in this ANOVA model. Average monthly soil moisture and Ψ stem were analyzed using repeated measures two-factor split-plot ANOVA to test the effects of Mulch, Irrigation and the Irrigation × Mulch interaction, with each combination of year and site analyzed separately, because Ψ stem and average monthly soil moisture content involved repeated measurements within each year that were not taken on the same days. Mulch and Irrigation were fixed effects, and Block was considered a random effect in the model. The multiple dates of Ψ stem measurement and months (average soil moisture content) within each year were treated as the repeated measures. Postharvest deficit irrigation was implemented after fruit harvest at Sites 2 and 3 in both 2016 and 2017, and at Site 1 in 2017. Therefore, it was only possible to analyze effects of Irrigation and the Irrigation ×Mulch interaction on fruit quality using data from Sites 2 and 3 in 2017; this ANOVA tested for an effect of 2016 PDI on fruit harvested the following year; i.e. in 2017. A repeated measures blocked one-factor ANOVA was performed to test the effect of Mulch on all fruit quality data and included data from Sites 2 and 3 in 2016 and 2017. Fruit quality data were also subjected to a blocked one-factor ANOVA testing the effect of Mulch that included data from all three sites in 2017 only. For all analyses, normality and homogeneity of variance were evaluated by examining histograms and qq plots of the residuals. Variables were log-transformed if they were of unequal variance or not normal. After data were transformed, each assumption for each test was examined again. All tests and test assumptions were performed using SAS [version 9.4. Cary, NC, USA]. Differences of LSMEANS calculated at P = 0.05 were used to identify the significance of differences among means.

Climatic conditions and soil moisture content
According to data from local Environment Canada weather stations near our sites in the north (Sites 1 and 2) and central (Site 3) Okanagan Valley (Environment Canada, 2021), 2017 was considerably drier than the other two years at all three of the sweet cherry orchards (Table S. The response of soil moisture to Irrigation is presented only for Sites 2 and 3 in 2017. After PDI was initiated in August 2017, it did not affect average monthly volumetric soil moisture compared to full irrigation for the remainder of the growing season at either site (Table S.5); however, the mulch treatments did have an effect. There was significantly higher soil moisture in bare and compostamended soils compared to woodchip-amended soils at Site 2 (Table S.5).

Soil chemical properties
According to the initial soil chemical analysis, all of the tested soil chemical properties were higher at the two newly established sites (Sites 1 and 2) than Site 3, except soil C: N ratio, which was higher at Site 3 than Sites 1 and 2 (Table S.2). The ANOVAs examining the Irrigation × Mulch interaction over two years at the two sites where PDI commenced in 2016 did not find significant effects of either Irrigation or the Irrigation × Mulch interaction for any soil chemical properties (Tables S.6, S.7). There was, however, a strong main factor effect of Mulch for most soil chemical properties in these twofactor ANOVAs, as well as the one-way ANOVAs encompassing all three sites and years (Tables 1, 2). The concentration of POXC, total organic C, organic matter, total C, total N, P, K, Mg, Na, pH, and EC were all higher in compost-amended compared to bare soils. Application of woodchips increased the total C and C: N ratio, but decreased soil Ca, extractable nitrate and ammonium, and EC relative to bare soils. Furthermore, the effect of mulch was influenced by year. Although the mulch effect spanned years, its intensity varied by year for all of the soil chemical properties measured (Tables 1,  2). The variables P, K, N and C were higher in 2015 under the compost treatment than the other years, whereas organic matter, Ca, Na and CEC were higher in 2017 under the compost treatment than the previous years (data not shown).

Foliar nutrient content and tree growth
Irrigation did not affect trunk cross-sectional area at Sites 2 and 3 in 2016 and 2017 (Table S.8); nor was there an Irrigation × Mulch interaction. Since sample leaves were collected after fruit harvest in the two growing seasons, but before PDI was implemented, it was only possible to assess the effects of Irrigation and the Irrigation × Mulch interaction on leaf characteristics in 2017 across Sites 2 and 3. Neither leaf area nor any of the foliar nutrients was affected by Irrigation or showed an Irrigation × Mulch interaction (Table S.8). Across the three sites and three growing seasons, Mulch affected all foliar nutrients except B (Table 3). Trees grown in compost-amended soils had higher foliar N, P, and K, and lower Mg and Ca concentrations than those grown in bare soils. Application of woodchips increased both foliar P and Mn concentrations, but reduced Cu concentrations compared to those from bare soils. Trees in woodchip-amended plots tended to have larger trunk cross-sectional areas (TCSA) relative to compost-amended and bare soils (Table 3). There was a significant Mulch × Year interaction over three sites and three years revealed for foliar K, Mg, Ca, and Mn; however, upon further analysis, there were no year to year differences in mulch treatment that were noteworthy (data not shown).

Stem water potential
The stem water potential (Ψ stem ) values across the three sites through the 2016 and 2017 growing seasons ranged from −2.40 to −0.34 MPa (Figure 1, 2). Stem water potentials were lowest (highly negative) in late August in 2016 at Site 3 ( Figure 1). In 2017, the Ψ stem generally decreased as the summer progressed, with the lowest values occurring in September at Sites 1 and 2 and in early August at Site 3 ( Figure 2). Consistent with the rather limited effects on soil moisture content (see 3.1), when each site was evaluated individually by year during which PDI was applied, there was no overall effect of PDI on stem water potential (Table S.9). There was a significant Irrigation × Mulch interaction at Site 2 in 2017 (Table S.9), with higher stem water potentials (less negative) in fully irrigated, woodchip-amended plots relative to bare soils exposed to PDI (data not shown).

Fruit quality and yield
Postharvest deficit irrigation was implemented after fruit harvest at Sites 2 and 3 in both 2016 and 2017, and at Site 1 in 2017. There was no effect of the 2016 post-harvest Irrigation treatments and no Irrigation × Mulch interaction on the quality of fruit harvested from Sites 2 and 3 the following year (i.e. 2017; data not shown). By 2017, cherries had been harvested from all three sites. In that year, across all three sites, firmness, colour, pH, and stem pull force were all affected by mulch treatment (Table 4). Compared to bare and woodchip treatments, compost treatment produced fruit with higher pH and stem pull force, both of which are positive attributes. Application of either compost or woodchips increased fruit colour relative to bare soils, whereas fruit firmness was greater under the compost treatment than the woodchip treatment. There was no effect of mulch treatment on total yield, SSC, or TA. Over 2016 and 2017, at the two sites (Sites 2 and 3) harvested in both those years, there was a greater number of splits in the woodchip treatment than in the other two treatments (Table S.10). Furthermore, fruit firmness, SSC, pH, TA, and yield were all higher in 2017 compared to 2016.

Irrigation × Mulch interaction effects
The main purpose of this study was to extend current knowledge of the effects of mulches on soil physicochemical properties, tree growth and fruit production, and to test whether mulch moderates the effects of PDI on these variables in a semi-arid region. For example, we hypothesized that the reduction in soil moisture due to PDI would be smaller under mulch than in bare soils, which would manifest in a significant Irrigation × Mulch term in the statistical analyses. However, the only significant Irrigation × Mulch interaction was for stem water potentials, which were greater (less negative) in the fully irrigated, woodchip treatment relative to the bare treatment exposed to PDI.   Table 3. Foliar macro and micronutrients, trunk cross-sectional area (TCSA), and leaf area as affected by mulch treatment at the three sites over three years. Data expressed as the mean of n = 12 plots per treatment per site and standard error (SE). Mulches or years sharing the same letter within a column do not differ significantly (P ≤ 0.05).  This result is indicative of additive benefits of mulch and full irrigation rather than moderation of the effect of PDI. We speculate that if the PDI treatment had been more extreme, i.e. more than a 24-28% reduction in irrigation water, then the moderating influence of mulch may have been more evident. Due to the lack of any other Irrigation × Mulch interactions in this study, the remainder of the Discussion will focus on the effects of Irrigation and Mulch, separately.

Irrigation effects
Postharvest deficit irrigation reduces water use and, depending on the level of deficit, has been demonstrated to affect the growth and productivity of different fruit tree crops (e.g. sweet cherry, peach and plum) (   moisture content has been used as a soil water status indicator for regulated deficit irrigation in sweet cherry and peach orchards (Dangi et al. 2016;Blanco et al. 2018). In our study, PDI did not affect average monthly volumetric soil moisture compared to full irrigation in 2017, the only year that the effect of PDI and full irrigation on soil moisture was compared, even though this was an unusually dry year. This result coincided with relatively high stem water potential values (i.e. less negative) in late September, regardless of irrigation treatment. Stem water potential is a reliable and sensitive physiological indicator of the water status in prune, plum, and cherry trees (Remorini and Massai 2003;Jones 2007;McCutchan and Shackel 2019). Most fruit trees exhibit stem water potential values ranging from −0.5 to −1.0 MPa under frequent irrigation (Shackel et al. 1977;McCutchan and Shackel 2019). In our study, most stem water potential values were within this range, except during the hottest months, particularly late August in 2016 at Site 3 and late September in 2017 at Site 2, when stem water potential values dropped to −2.0 and −1.7 MPa, respectively. The limited effect of PDI on soil moisture and stem water potential indicates that sufficient water was supplied, even with the reduction in irrigation, to meet plant water needs across sites. There were no differences detected in any soil chemical properties between PDI and full irrigation. Postharvest deficit irrigation with a 24-28% reduction in applied water did not affect cherry foliar nutrient concentration, tree growth (as measured by TCSA), productivity (leaf area), or fruit yield and quality when compared to full irrigation. In line with our results, Abrisqueta et al. (2011) found that foliar nutrient composition remained similar after implementation of 24-50% reduced irrigation at different growth stages for peach trees in Spain, and similar vegetative growth was achieved with full irrigation and the deficit treatment (24% reduced) by apricot trees in a three-year study in Turkey (Bozkurt et al. 2015). Postharvest deficit irrigation is applied after harvest when trees are least sensitive to water stress, which may explain why many reports, including ours, have not found that fruit yield or quality is negatively affected by PDI (Goldhamer and Beede 2004;Marsal et al. 2010;Behboudian et al. 2011;Chai et al. 2014). Nevertheless, it is important to note that more extreme levels of PDI would likely have negative effects on cherry trees, and such effects are likely to vary by cultivar-rootstock combination, soil type and climatic conditions (Neilsen and Kappel 1996;Galindo et al. 2018). The total amount of water applied could only be determined for 2017, which was also the only year that the PDI treatment was imposed at all three sites, and was roughly estimated to be 425,304 L at Site 1, 312,768 L at Site 2 and, 933,120 L at Site 3. By reducing applied water in the PDI treatment by 26, 24, and 28% at Sites 1, 2, and 3, respectively, relative to full irrigation, the estimated amount of water saved in 2017 was 424,044 L at Site 1, 304,704 L at Site 2, and 843,264 L at Site 3, which is 1,572,012 L in total. If PDI had not been imposed, approximately 48.9% more water would have been applied cumulatively across the three sites.

Mulch effects
In this study, we found little effect of Mulch treatments on soil moisture content. An exception was at Site 2, where there was lower soil moisture content in woodchip-treated soils than in bare or compost-treated soils in 2017. This decrease in soil moisture under woodchip mulch was unexpected as most previous studies have shown that the application of various types of mulch (e.g. straw, plastic film) prevents soil water evaporation and retains soil moisture (Kumar and Dey 2011;Kasirajan and Ngouajio 2012;Kader et al. 2017;Jones et al. 2020). The mulch itself could have absorbed much of the irrigation water during each irrigation cycle, reducing the amount of water entering the mineral soil, or the mulch could have developed a hydrophobic surface (Finkenstadt and Tisserat 2010), causing some of the irrigation water to run off to the side of the tree row where soil moisture was being measured. Because mulches had no or little effect on soil moisture in this study, it is not surprising that no effect on tree stem water potential was observed. Yard waste-based composts are rich sources of organic C and several plant nutrients, such as N, P and K. We found that within one year of application the compost mulch resulted in increased concentrations of total C, POXC, N, P, K, Na and Mg in underlying soil. These nutrient inputs also resulted in increased soil pH and EC, consistent with previous studies with yard waste composts (Milošević et al. 2013;Safaei Khorram et al. 2019;Thompson et al. 2019). In this study, woodchip mulch also tended to have a positive effect on total nutrient contents in underlying soil when compared to the bare ground but these effects were smaller than those of the compost mulch. In our study, trees that were grown in composttreated soil had higher foliar N, P and K, and lower Mg and Ca levels than those grown on bare ground, while those grown with woodchips had higher P and Mn concentrations. Higher levels of foliar N, K, and P from compost or woodchip treatments were likely due to the increased availability of these nutrients in treated soils, while lower Mg and Ca may have resulted from displacement by K and effects of higher pH (Yike Bing 2011; Al et al. 2018). Although soil biological parameters are not presented here, it is worth noting that previous studies have demonstrated that compost and wood chip mulches can increase indicators of soil biological activity and suppress plant-parasites , Forge et al. 2008Watson et al. 2017). Woodchip mulch substantially reduced concentrations of available forms of N in underlying soil. This effect was most likely the result of microbial immobilization resulting from the very high C:N ratio of the woodchip material (Gonzalez and Cooperband 2002;Yang et al. 2003;Gale et al. 2006). Despite the immobilization of nitrate and ammonium in soil, woodchip mulch did not suppress leaf N concentrations relative to bare soil, and leaf N concentrations were all within the range considered adequate (BC Tree Fruit Production Guide 2021), indicating that the growers' fertilization practices compensated for N immobilization by the woodchip mulch. We expect that with time, as the woodchip mulch decomposes and the C: N ratio becomes narrower, the strong N immobilization in underlying soil will subside. Although fruit yield was not affected by mulch treatment, some aspects of fruit quality including pH, stem pull force, and color were improved by the application of compost mulch. We also detected an increase in fruit color and a decrease in fruit firmness from woodchip mulch. Some of our fruit quality results are in contrast to a recent report on sweet cherry in Australia, which showed that a humified compost made from manure, amended for two growing seasons, increased total soluble solids and decreased firmness and stem pull force, without influencing export quality (Tan et al. 2018). The differences between our results and previously reported findings could be due to variations in cultivated varieties and the type of organic materials used (e.g. mulch types). Additionally, in partial agreement with our findings, the use of bark mulch improved the fruit quality of sweet cherry during the first few seasons (Neilsen et al. 2006). These increased fruit quality attributes of fruit trees grown under compost and woodchip mulch could be due to the presence of higher levels of organic materialderived soil nutrients, particularly P and K, which can improve fruit quality (e.g. color and total soluble solids) (Štampar et al. 2015;Hasanuzzaman et al. 2018).

Conclusions
Post-harvest deficit irrigation, with a 24-28% reduction in the volume of water applied, had almost no effects on soil chemical properties, tree growth or crop yield and quality in three sweet cherry orchards in the Okanagan Valley. In 2017, the estimated water applied ranged from 312,000 to 933,000 L and the estimated water saved with the application of PDI ranged from 304,000 to 843,000 L, amounting to a total savings of 48.9% of water across the three sites. This suggests that it is a viable approach to conserve water use while maintaining fruit production in cherry orchards in this semi-arid region. Independent of irrigation rate, surface application of compost improved soil physicochemical properties, specifically by increasing organic matter and major soil nutrients, and both mulches increased plant macronutrient concentrations in foliage. Although our study only spanned three years, with data only available for one or two years for some variables, there were early signs that mulches could increase tree growth and some fruit quality attributes. Longer term studies would facilitate our understanding of changes over time, as well as year-to-year variability in perennial crop systems. Overall, our results indicate that the use of organic mulches that are applied annually in newly planted sweet cherry orchards may be an effective tool to improve and/ or restore soil organic matter and nutrients, maintain soil fertility, and improve plant growth and productivity in the long term. Overall, this study showed that PDI and organic mulches could be effective management strategies that conserve water and soil resources and more sustainably produce sweet cherry in newly planted sweet cherry orchards in the Okanagan Valley of British Columbia, Canada, with no compromise in tree growth or fruit yield and quality. The results of this study have potential applicability to sustainable tree fruit production in other semi-arid regions.