Next Article in Journal
Automatic Sun Glint Removal of Multispectral High-Resolution Worldview-2 Imagery for Retrieving Coastal Shallow Water Parameters
Next Article in Special Issue
A Combination of Plant NDVI and LiDAR Measurements Improve the Estimation of Pasture Biomass in Tall Fescue (Festuca arundinacea var. Fletcher)
Previous Article in Journal
Estimating Soil Moisture with Landsat Data and Its Application in Extracting the Spatial Distribution of Winter Flooded Paddies
Previous Article in Special Issue
Correction: Tilly, N. et al. Fusion of Plant Height and Vegetation Indices for the Estimation of Barley Biomass. Remote Sens. 2015, 7, 11449–11480
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Airborne Thermal Imagery to Detect the Seasonal Evolution of Crop Water Status in Peach, Nectarine and Saturn Peach Orchards

1
Efficient Use of Water Program, Institut de Recerca i Tecnologia Agroalimentàries (IRTA), Fruitcentre, Parc Cientific i Tecnològic Agroalimentari (PCiTAL), Lleida 25003, Spain
2
Center of Spatial Technologies (CSTARS), Department of Land, Air and Water Resources (LAWR), University of California, Davis, CA 95616, USA
3
Institute for Sustainable Agriculture (IAS), Consejo Superior de Investigaciones Científicas (CSIC), Alameda del Obispo, s/n, Córdoba 14004, Spain
*
Author to whom correspondence should be addressed.
Remote Sens. 2016, 8(1), 39; https://doi.org/10.3390/rs8010039
Submission received: 7 October 2015 / Revised: 8 December 2015 / Accepted: 29 December 2015 / Published: 5 January 2016
(This article belongs to the Special Issue Remote Sensing in Precision Agriculture)

Abstract

:
In the current scenario of worldwide limited water supplies, conserving water is a major concern in agricultural areas. Characterizing within-orchard spatial heterogeneity in water requirements would assist in improving irrigation water use efficiency and conserve water. The crop water stress index (CWSI) has been successfully used as a crop water status indicator in several fruit tree species. In this study, the CWSI was developed in three Prunus persica L. cultivars at different phenological stages of the 2012 to 2014 growing seasons, using canopy temperature measurements of well-watered trees. The CWSI was then remotely estimated using high-resolution thermal imagery acquired from an airborne platform and related to leaf water potential (ѰL) throughout the season. The feasibility of mapping within-orchard spatial variability of ѰL from thermal imagery was also explored. Results indicated that CWSI can be calculated using a common non-water-stressed baseline (NWSB), upper and lower limits for the entire growing season and for the three studied cultivars. Nevertheless, a phenological effect was detected in the CWSI vs. ѰL relationships. For a specific given CWSI value, ѰL was more negative as the crop developed. This different seasonal response followed the same trend for the three studied cultivars. The approach presented in this study demonstrated that CWSI is a feasible method to assess the spatial variability of tree water status in heterogeneous orchards, and to derive ѰL maps throughout a complete growing season. A sensitivity analysis of varying pixel size showed that a pixel size of 0.8 m or less was needed for precise ѰL mapping of peach and nectarine orchards with a tree crown area between 3.0 to 5.0 m2.

Graphical Abstract

1. Introduction

Water scarcity is a major concern in many irrigated agricultural areas of the world. Therefore, irrigation water management needs to be carried out more efficiently, aiming at saving water and maximizing its productivity [1]. The adoption of regulated deficit irrigation (RDI) techniques, using water stress indicators, has been used in many fruit trees in order to optimize water use without affecting crop yields. In peach trees (Prunus persica L.), the benefit is well-known for applying RDI during stage II of fruit development. Besides saving water, RDI has the potential to reduce excessive tree vigor, maintain yield, and may improve fruit quality [2,3]. However, the difficulty in identifying spatial variability of the crop water status across orchards limits its practical implementation in commercial fields. The success of RDI techniques strongly depends on the appropriate use of on-farm irrigation system and on the capacity to use new tools to detect crop water status across orchards.
Remote sensing technologies are a successful tool to accurately identify spatial heterogeneity and have been shown to have the capacity to monitor irrigation at orchard level [4,5]. The use of remote sensing in the assessment of crop water status through canopy temperature has had a long development history. In the 1970s and 1980s, canopy temperature was first suggested as a method to detect stress using hand-held thermal infrared thermometers [6,7,8,9,10,11,12]. However, these measurements were site-specific and not useful to assess spatial distribution of water status within a field. The advent of modern remote sensing technologies offers the possibility to develop temperature-derived indicators from airborne thermal imaging and map spatial variability of water status.
Working on herbaceous crops, Jackson et al. [11] and Idso et al. [8] developed the concept of crop water stress index (CWSI) as a thermally base stress indicator. Since then, the CWSI has been widely used in annual crops [13,14,15], and more recently in perennial crops such as grapevines [16,17], olive [18,19], pistachio [20], citrus [21] and peach [22]. CWSI is inversely related to transpiration rate and stomatal conductance [7] and has been successfully related to indicators of crop water status such as leaf water potential (ѰL) [23,24]. However, stomatal control of canopy conductance and ѰL are highly sensitive to VPD, and change between cultivars and throughout crop development [25,26]. Such changes may affect the relationship between CWSI and ѰL. In fact, Bellvert et al. [27] reported significant differences in the CWSI - ѰL relationships between grapevine cultivars and phenological stages.
To use remotely estimates of leaf water potential (Ѱest) for assessing the spatial variability in water stress across peach orchards, it would be necessary to find out the seasonal variations in the CWSI vs. ѰL relationship. The present study tested the hypothesis that Ѱest is a reliable water stress indicator in different Prunus persica L. cultivars throughout different growing seasons, and whether there are different responses between cultivars. To achieve it, “non-water-stressed baselines”, which are necessary for calculating CWSI, were developed over two seasons, and the CWSI was seasonally related with ѰL. The feasibility of mapping within-orchard spatial variability of Ѱest from high-resolution thermal imagery, remotely acquired by an airborne platform was also explored.

2. Materials and Methods

2.1. Study Site

The study was carried out from 2012 to 2014 growing seasons in three commercial orchards located near Lleida, Spain. In 2012 and 2013, the study was carried out in a 2-ha peach (41°41′64″N, 0°32′91″E) orchard planted in 1996 with an early maturing cultivar of peach (Prunus persica cv. Royal Glory) in a 4 m × 3 m grid. In the same two years, results obtained in the peach orchard were validated in a 15 year old 2.2-ha nectarine orchard (Prunus persica cv. Big Top) planted in a 4 m × 3 m grid (41°41′72″N, 0°32′87″E). In addition, results for peach were validated in 2014 in a 3 year old 2.8-ha Saturn peach orchard (Prunus persica cv. Platicarpa) with tree spacing of 4 × 1.5 m grid (41°41′64″N, 0°32′38″E). Tree crown area ranged from 3.0 to 5.0 m2 for peach and nectarine and from 1.5 to 2.5 m2 for Saturn peach. The climate of the area is Mediterranean, with an average annual rainfall and evapotranspiration of 364 and 1088 mm, respectively. Within each orchard, three irrigation treatments were set up to test different levels of water status throughout the season. The treatments were: (i) full-irrigation control, where irrigation replaced 100% ETc; (ii) moderate deficit irrigation, where irrigation replaced 50% ETc; and (iii) deficit irrigation, where water was applied only after midday leaf water potential dropped below −2.0 MPa. Each irrigation treatment consisted of four adjacent rows with eight trees per row. The trees of the two central rows were monitored while the others served as a border. Irrigation water was applied through a drip irrigation system with emitters spaced 0.75 m apart on a single drip line per row, discharging 8 L·h−1. Trees were irrigated daily and scheduled on a weekly basis. Water requirements were calculated using a water balance technique to replace crop evapotranspiration (ETc) as follows: ETc = (ETo × Kc)—rainfall. ETo and Kc represent the reference evapotranspiration and crop coefficient respectively. The Penmann-Monteith method was used to determine ETo (Allen et al., [28]) and Kc vales were obtained from Doorenbos and Pruitt [29].

2.2. Infrared Temperature Data

Four infrared temperature sensors (IRTS) (model PC151LT-0; Pyrocouple series, Calex Electronics Limited, Bedfordshire, UK) were installed about 1.5 m above two peach trees of the full-irrigated control treatment (two sensors in each tree). Canopy temperature was measured from beginning May to mid-September for 2012 and 2013, obtaining data throughout the growing season. The calibrated IRTS were installed aiming vertically downward (nadir view). The angular field of view was 15:1 with an accuracy of ±1%, which measured a canopy target spot around 0.12 m diameter. Visual inspection of sensor positioning, comparison with air temperature (Ta) data and consistency with values measured with a hand-held infrared thermometer (Fluke 62 mini, Fluke Europe, Eindhoven, The Netherlands) ensured that 100% of the temperature signal came only from the leaves. All IRTS were connected to a datalogger (model CR200X; Campbell Scientific, Logan, UK) that recorded temperatures every minute and stored the 15-min averages. Recorded data of full-irrigated control trees was used to calculate the baselines of the crop water stress index (CWSI). The empirical CWSI was calculated as Idso et al. [9]:
C W S I = ( T c T a ) ( T c T a ) L L ( T c T a ) U L ( T c T a ) L L
where Tc-Ta is measured canopy-air temperature difference; (Tc-Ta)LL lower limit of (Tc-Ta) of a canopy which is transpiring at the potential rate, and (Tc-Ta)UL expected differential in the case of a non-transpiring canopy. (Tc-Ta)LL is function of vapour pressure deficit (VPD) (non-water-stressed baseline, NWSB). The NWSB was calculated in peach trees for two years following the procedure described in Bellvert et al. [17]. Only data of sunny days, which had wind speed below 6 m·s−1 (at a height of 10 m) were used in the assessment of CWSI. Hourly values of (Tc-Ta) were regressed against vapour pressure deficits (VPD) separately for the hours of the day, from 10:00 to 16:00 hours (solar time). Following recommendations of Bellvert et al. [17,27], the linear NWSBs were transformed to curvilinear equations to obtain the upper limits. The upper limits (UL) were obtained by solving the NWSB curvilinear equation for VPD = 0 and then correcting for vapor pressure differences caused by Tc-Ta [9]. The physical theory assigns a straight horizontal line to UL, which is irresponsive to VPD because this is only thought as dependent on relative humidity (HR). However, when different days are compared in a semi-arid climate, the VPD tends to be higher during days with high air temperature (Ta). Therefore, days with higher Ta have had higher VPD (Table 1). The UL increase with VPD reflects the effect of higher Ta on leaf energy balance. The lower limit was obtained taking the minimum values of Tc-Ta for each VPD. The NWSBs were obtained for each phenological stage: (1) stage II—lignification of endocarp (pit hardening); (2) stage III—maturation; (3) Post-harvest—from harvest (beginning July) to end of August. Stage I (considered in this study to extend from fruit set to initial pit hardening) was not studied because of low canopy cover. Pooled data from 2012 and 2013 was used to develop a common NWSB, lower and upper limits for each phenological stage and for the entire growing seasons. Air temperature (Ta), and vapor pressure deficit (VPD) were obtained from a portable weather station (Watchdog, model 2900ET, Spectrum Technologies, Inc., Plainfield, IL, USA) located on one side of the orchard, and used to obtain local meteorological conditions throughout the season.
Table 1. Climate data at the commercial orchards during the flight time.
Table 1. Climate data at the commercial orchards during the flight time.
YearFlight DayStageOrchardTair (°C)VPD (kPa)Rad (W·m−2)Wind Speed (kph)
201214 MayIIP + N24.02.49266
1 JuneIIP + N31.33.89477
15 JuneIIIP + N32.13.49598
6 JulyIIIP + N27.62.59338
31 JulyPHP + N31.42.991410
21 AugustPHP + N34.04.28497
20137 MayIIP + N24.41.690411
13 MayIIP + N20.01.610524
4 JuneIIP + N23.61.610836
11 JuneIIIP + N28.52.610823
25 JuneIIIP + N22.51.610788
3 JulyIIIP + N29.62.388212
16 JulyPHP + N33.03.010323
31 JulyPHP + N33.73.810438
12 AugustPHP + N31.33.39683
201430 AprilIISP21.91.690216
13 MayIISP21.11.8105025
29 MayIISP19.71.09604
6 JuneIIISP28.21.910349
13 JuneIIISP30.82.39400
27 JuneIIISP28.22.611840
27 AugustPHSP30.22.396310
P: Peach, N: Nectarine, SP: Saturn peach.

2.3. Airborne Campaign

Flights were conducted with a thermal sensor (FLIR SC655, FLIR Systems, Wilsonville, OR, USA) installed on an aircraft (CESSNA C172S EC-JYN). The camera had a resolution of 640 × 480 pixels, equipped with a 13.1 mm optics focal length yielding an angular FOV of 45° that delivered approximate ground resolution of 0.15 to 0.20 m. The spectral response was in the range of 7.5–13 μm. The radiometric calibration of the sensor was assessed in the laboratory using a blackbody (model P80P, Land Instruments, Dronfield, UK). In addition, various calibrations were conducted at the time of each flight using surface temperature measurements to improve the calibration. The accuracy of this method is discussed in Berni et al. [19,30], who demonstrated an accuracy error less than 1 K using a similar camera on board an Unmanned Aerial Vehicle (UAV). The sensor was connected to a computer via USB 2.0 protocol. The software to acquire thermal images was developed at the Laboratory for Research Methods in Quantitative Remote Sensing (Quantala, IAS-CSIC, Córdoba, Spain) and described in Zarco-Tejada et al. [31].
The peach and nectarine orchard study sites used for field data collection can be seen in Figure 1a. The flights were conducted at 12:00 solar time (14:00 local time) at 150 m altitude above the ground level. The high-resolution thermal images enabled discrimination of pure tree crowns from soil background effects (Figure 1b). The flying pattern consisted of four longitudinal lines of 1500 m separated by 65 m. Table 1 lists imagery acquisition dates, phenological stages and weather conditions for each flight throughout the growing season. Concomitant with each flight, leaf water potential (ΨL) was measured in 18 trees in each orchard (six per irrigation treatment). Two fully expanded leaves exposed to direct sunlight were measured on each tree. A Scholander pressure chamber (Soil Moisture Equipment Corp., Santa Barbara, CA, USA) was used following the recommendations of Turner and Long [32]. Aluminum foil was used between rows as a target to mark the exact location of measured trees. Mean crown temperature (Tc) was extracted from six trees per irrigation treatment and used to calculate the crop water stress index (Figure 1c).
Figure 1. Thermal mosaic acquired at 0.20 m pixel resolution, observing in (a) nectarine and peach orchards. In red are shown the areas where the irrigation treatments were located; (b) detailed study site area used for field data collection in the peach orchard. Trees under different irrigation treatments (full-irrigation control, moderate and deficit irrigation) were measured and located with aluminium foil between rows; and (c) peach trees under different irrigation treatments used for pure tree crown temperature extraction.
Figure 1. Thermal mosaic acquired at 0.20 m pixel resolution, observing in (a) nectarine and peach orchards. In red are shown the areas where the irrigation treatments were located; (b) detailed study site area used for field data collection in the peach orchard. Trees under different irrigation treatments (full-irrigation control, moderate and deficit irrigation) were measured and located with aluminium foil between rows; and (c) peach trees under different irrigation treatments used for pure tree crown temperature extraction.
Remotesensing 08 00039 g001
Simultaneously with each flight in 2013, stomatal conductance (gs) was also measured in the nectarine orchard using a steady-state porometer (model LI-1600, Li-Cor Inc., Lincoln, NE, USA). gs was measured in the same leaves where ΨL was recorded.

2.4. Spatial Resolution Assessment

The influence of pixel size on crop water status detection was studied for determining the minimum and best possible spatial resolution that may be required in assessing Tc of peach and nectarine trees. The thermal imagery acquired on 1 and 15 June 2012 were re-sampled using a pixel aggregate technique through cubic convolution. Newly obtained re-sampled images had pixel sizes of 0.3, 0.6, 1.0, 1.2, 1.5, and 2.0 m. Exactly the same region of interest created for the eighteen peach and nectarine trees for the very high resolution thermal imaging was used to extract the aggregated pixels from the lower resolution mosaics.

2.5. Estimation of ѰL from CWSI and Validations

Estimation of leaf water potential for peach trees was performed comparing two methods: (1) using CWSI for each specific phenological stage; and (2) using a common CWSI pooling data from two consecutive growing seasons. In addition, the methodology proposed to estimate ΨL in peach trees was also validated for nectarine and Saturn peach trees in different years. The “non-water-stressed baselines” developed for peach trees were then also used to calculate CWSI for these two cultivars.

2.6. Statistical Analysis

Non-watered-stressed baselines (NWSB) were transformed to a linear regression model for the purpose of analyzing differences between phenological stages. A covariance analysis (ANCOVA) was performed to analyze differences between years and phenological stages using the SAS statistical package [33]. Specific differences among phenological stages in the slopes and intercept of the lines were subsequently tested by orthogonal contrasts.

3. Results

3.1. Non-Water-Stressed Baselines

The relationship between difference of canopy and air temperature (Tc-Ta) with vapor pressure deficit (VPD) for well-watered peach trees was significant for the two years of study (Figure 2). The Tc-Ta trend decreased as VPD increased. The coefficients of determination (R2) ranged from 0.61 to 0.85. Although not significant differences were detected between years during stages II and III, the differences in the post-harvest stage were significant. During post-harvest, the intercept in the Tc-Ta vs. VPD relationship was significantly higher in 2012 than in 2013. However, a common baseline using the two years of data was performed for each phenological stage. The statistical analysis indicated a different phenological response within years (Table 2). The intercept of the relationship Tc-Ta vs. VPD was significantly different for two studied years (p < 0.0001). In 2012, Tc-Ta vs. VPD relationship from stage II presented a higher intercept than stage III. On the other hand, the intercept of post-harvest stage in 2013 was significantly lower in comparison with earlier stages. In the same year, the phenological stage III had the lowest slope and post-harvest had the highest.
Table 2. A covariance analysis (ANCOVA) analysis of Tc-Ta for phenological stages at different years and probabilities tested by orthogonal contrasts of slopes (Stage × vapour pressure deficit (VPD)) and intercepts (Stage).
Table 2. A covariance analysis (ANCOVA) analysis of Tc-Ta for phenological stages at different years and probabilities tested by orthogonal contrasts of slopes (Stage × vapour pressure deficit (VPD)) and intercepts (Stage).
20122013All Years
VPD<0.0001 *<0.0001 *<0.0001 *
Stage<0.0001 *<0.0001 *<0.0001 *
Stage × VPD0.957<0.0001 *0.0641
Contrast **StageStageStage × VPDStage
Stage II vs. III0.0414-0.04510.0197
Stage II vs. PH-0.0059<0.00010.0431
Stage III vs. PH-<0.0001<0.0001-
* Significance at p < 0.0001 (SAS 2002); ** Significance at p < 0.05 (SAS 2002).
The Tc-Ta vs. VPD relationships were used to develop the curvilinear NWSBs, and lower and upper limits for each phenological stage and for the entire growing season (Figure 3). It is interesting to emphasize that LL and UL of stage III had a slightly different slope and intercept in comparison with other stages, perhaps related to a higher Ta and global solar radiation during that period. These differences were considered negligible and a common LL and UL was used for the entire growing season (Figure 3c).
Figure 2. Relationships between the difference of canopy and air temperature (Tc-Ta) and vapour pressure deficit (VPD) for the 10:00 to 16:00 hour data of fully irrigated control peach trees, showing: (ac) differences between years (2012 and 2013) at different phenological stages (Stages II and III, and post-harvest); and (d) seasonal response for two years data. All relationships were significant (p < 0.0001).
Figure 2. Relationships between the difference of canopy and air temperature (Tc-Ta) and vapour pressure deficit (VPD) for the 10:00 to 16:00 hour data of fully irrigated control peach trees, showing: (ac) differences between years (2012 and 2013) at different phenological stages (Stages II and III, and post-harvest); and (d) seasonal response for two years data. All relationships were significant (p < 0.0001).
Remotesensing 08 00039 g002
Figure 3. (a) Non-water-stressed baselines (NWSB), and (b) lower and upper limits, used to calculate the CWSI for peach trees at different phenological stages (Stages II and III, Post-harvest) and (c) for the entire growing season (All Season). All relationships were significant (p < 0.0001).
Figure 3. (a) Non-water-stressed baselines (NWSB), and (b) lower and upper limits, used to calculate the CWSI for peach trees at different phenological stages (Stages II and III, Post-harvest) and (c) for the entire growing season (All Season). All relationships were significant (p < 0.0001).
Remotesensing 08 00039 g003

3.2. Relationship between CWSI and ΨL for Peach Trees

Midday ΨL correlated significantly with CWSI for the entire growing season and followed a curvilinear model during pre-harvest stages and a linear model during post-harvest (Figure 4). This means that during pre-harvest, ΨL started to decline significantly only after a threshold CWSI value. A significant seasonal effect was also observed in this relationship, indicating that for a particular level of CWSI, the values of ΨL were lower as the crop developed.
Figure 4. (ac) Comparison of the relationship between crop water stress index (CWSI) and leaf water potential (ΨL) for peach trees at different phenological stages of the 2012 and 2013 growing seasons. CWSI has been calculated using the lower and upper limits for each specific phenological stage (Stages) and for all the season (All Season). All relationships were significant (p < 0.0001).
Figure 4. (ac) Comparison of the relationship between crop water stress index (CWSI) and leaf water potential (ΨL) for peach trees at different phenological stages of the 2012 and 2013 growing seasons. CWSI has been calculated using the lower and upper limits for each specific phenological stage (Stages) and for all the season (All Season). All relationships were significant (p < 0.0001).
Remotesensing 08 00039 g004
CWSI was calculated under two different methods: (i) using baselines (lower and upper limits) for each phenological stage; and (ii) using a common baseline for the entire growing season. Visually, the two methods had similar responses without differences in the functions and coefficients of determination (R2) (Figure 4). In order to identify which method had the strongest influence on ΨL estimation, ΨL was estimated using both methods and related to the observed ΨL by linear regression. The one-to-one relationships between estimated and observed ΨL were significant for all phenological stages (p < 0.0001), and no significant differences were found between methods used for estimating ΨL (Figure 5). There were differences in the one-to-one relationships at different phenological stages. Despite the regression from stage II having the lowest root mean square error (RMSE), it also presented a low slope and high intercept. Stage III had a higher RMSE than stage II, but the intercept was closer to zero and the slope to one. Although intercept and slope of regressions from the post-harvest stage were similar to stage III, the RMSE was higher (Figure 5c,g). The best fit regression was obtained taking into account data from the entire season (Figure 5d,h). This comparison was done for peach trees using data from two consecutive years, and showed that the CWSI vs. ΨL relationships in 2012 agreed with those of 2013.

3.3. Validation for Nectarine and Saturn Peach

The algorithms of CWSI developed for peach trees were validated for nectarine and Saturn peach trees. The seasonal relationships between midday ΨL and CWSI for nectarine and Saturn peaches followed the same trend as those of peach trees (Figure 6). Coefficients of determination ranged from 0.52 to 0.78 and were similar to those presented in Figure 4 for peach trees. Estimates of leaf water potential using seasonal pooled data from all cultivars fitted significantly well with observed ΨL (p < 0.0001) (Figure 7). However, differences in their slopes and intercepts were detected in comparison with estimates of ΨL only using peach tree data. The regression at stage II had the lowest R2 and slope and a high intercept (Figure 7a). In Stage III, despite the RMSE was 0.20 MPa, the slope and intercept of the regression were respectively closer to one and zero in comparison with the same regression only with peach tree data (Figure 7b). The pertinent regressions from post-harvest and all season had a slight decrease in the adjusted one-to-one parameters in comparison with those presented in Figure 3 at their respective stages (Figure 7c,d). Overall, despite a slight decrease in the one-to-one regressions with data from all three cultivars in comparison with only data from peach trees, it seems that estimates of ΨL are robust throughout the season.
Figure 5. (ah) Relationships between observed and estimated ΨL for peach trees at different phenological stages of the 2012 and 2013 growing seasons. Estimated ΨL has been obtained from the calculation of CWSI using: (left) a common equation for all the season; and (right) equations for each phenological stage. All relationships were significant (p < 0.0001).
Figure 5. (ah) Relationships between observed and estimated ΨL for peach trees at different phenological stages of the 2012 and 2013 growing seasons. Estimated ΨL has been obtained from the calculation of CWSI using: (left) a common equation for all the season; and (right) equations for each phenological stage. All relationships were significant (p < 0.0001).
Remotesensing 08 00039 g005
Figure 6. (ac) Seasonal validation of the relationship between CWSI and leaf water potential (ΨL) in nectarine for the 2012 and 2013 growing season and Saturn peach trees for 2014 growing season. Data obtained for peach trees has been added as standard comparison. All relationships were significant (p < 0.0001).
Figure 6. (ac) Seasonal validation of the relationship between CWSI and leaf water potential (ΨL) in nectarine for the 2012 and 2013 growing season and Saturn peach trees for 2014 growing season. Data obtained for peach trees has been added as standard comparison. All relationships were significant (p < 0.0001).
Remotesensing 08 00039 g006
Figure 7. (ad) Relationships between observed and estimated ΨL for peach, nectarine and Saturn peaches at different phenological stages using data from 2012 to 2014 growing seasons. All relationships were significant (p < 0.0001).
Figure 7. (ad) Relationships between observed and estimated ΨL for peach, nectarine and Saturn peaches at different phenological stages using data from 2012 to 2014 growing seasons. All relationships were significant (p < 0.0001).
Remotesensing 08 00039 g007

3.4. Minimum Pixel Size to Detect Water Stress

High resolution thermal imagery enabled the retrieval of pure tree crown pixels, while lower spatial resolution has more chances to contain pixels with mixed information of canopy, shadows and soil background, making it difficult to remotely estimate ΨL with sufficient precision.
The relationship between ΨL and CWSI tend to reduce R2 with increasing pixel size (Table 3). Thermal mosaics with a pixel size of 0.15 and 0.30 m had the highest R2. Also, the regressions between measured and estimated ΨL for peaches and nectarines were well adjusted and had the lowest RMSE. Contrary to results provided by Bellvert et al. [17] in grapevines, which indicated that a minimum pixel size of 0.30 m was the optimum, it seems that for peach and nectarine trees, pixel size can be higher. Pixel sizes from 0.60 to 0.80 m almost had not a negative effect on R2, and the RMSE for those pixel sizes only increased around 6% in comparison with pixels at 0.30 m. On the other hand, despite pixels at 1.00 m pixel size having a significant relationship, the RMSE was too high. There was no significant relationship between ΨL and CWSI from 1.20 m resolution.
Table 3. Comparison of the coefficients of determination (R2) of the relationship between ΨL (y) and CWSI (x) for grapevines obtained by Bellvert et al. [17] and peach and nectarine trees at different spatial pixel resolutions ranging from 0.15 to 2.00 m. R2 for peach and nectarine trees corresponded with data from thermal images acquired on 1 and 15 June 2012. Ψest (y) from the ΨL vs. Ψest regression was estimated from the equation corresponding with stage III (Ψest = −1.171CWSI2 − 0.714CWSI − 0.805).
Table 3. Comparison of the coefficients of determination (R2) of the relationship between ΨL (y) and CWSI (x) for grapevines obtained by Bellvert et al. [17] and peach and nectarine trees at different spatial pixel resolutions ranging from 0.15 to 2.00 m. R2 for peach and nectarine trees corresponded with data from thermal images acquired on 1 and 15 June 2012. Ψest (y) from the ΨL vs. Ψest regression was estimated from the equation corresponding with stage III (Ψest = −1.171CWSI2 − 0.714CWSI − 0.805).
GrapevinesPeaches & Nectarines
ΨL vs. CWSIΨL vs. Ψest
Pixel Size (m)R2R2EquationEquationRMSE
0.15-0.76 *y = −1.708 x2 + 0.194x − 0.993y = 0.905x − 0.2630.21
0.300.710.75 *y = −1.669 x2 + 0.252x − 1.060y = 0.942x − 0.1870.21
0.600.380.66 *y = −0.573 x2 − 0.707x − 0.883y = 0.986x − 0.1470.27
0.800.270.65 *y = −0.249 x2 − 0.979x − 0.850y = 1.053x − 0.0440.28
1.000.220.56 *y = −0.224 x2 − 0.743x − 0.908y = 1.204x − 0.0090.46
1.20<0.100.28y = − 0.742x − 0.997y = 0.808x − 0.5860.56
1.500.280.11y = − 0.468x − 1.089y = 0.651x − 1.1430.88
2.000.29<0.10---
* Relationships were significant at p = 0.05 (SAS 2002).

3.5. Maps of Remotely Estimated ΨL

The information developed in this study enables the generation of maps of remotely sensed leaf water potential (Ψest) throughout a complete growing season. To carry this out successfully, the seasonal relationships reported in this study must be taken into account. Figure 8 shows Ψest maps of the peach and nectarine orchards at different phenological stages in the 2013 growing season. The farmer applied two different irrigation strategies in different irrigation sectors: (i) Control, full-irrigation throughout the growing season; and (ii) Regulated Deficit Irrigation (RDI), which consisted on applying mild deficit irrigation during stage II and severe deficit during post-harvest. During stage III, both irrigation sectors were fully irrigated. Ψest maps illustrated the spatial variability across the orchard and the differences in crop water status between irrigation sectors. For instance, those irrigation sectors under RDI had lower Ψest during stage II and post-harvest than the Controls.
Figure 8. Maps of remotely sensed leaf water potential (Ψest) at different phenological stages of the 2013 growing season for a 2-ha peach (ad) and 2.2-ha nectarine orchard (ef). Two different irrigation treatments were applied at irrigation sector level (full irrigation, CONTROL) and Regulated Deficit Irrigation (RDI). Averaged Ψest values for each irrigation sector are indicated. The equations applied to estimate Ψest from CWSI were the following: y = −1.846x2 + 0.426x − 0.773 (stage II), y = −1.171x2 − 0.714 − 0.805 (stage III), and y = −1.558x − 1.144 (post-harvest).
Figure 8. Maps of remotely sensed leaf water potential (Ψest) at different phenological stages of the 2013 growing season for a 2-ha peach (ad) and 2.2-ha nectarine orchard (ef). Two different irrigation treatments were applied at irrigation sector level (full irrigation, CONTROL) and Regulated Deficit Irrigation (RDI). Averaged Ψest values for each irrigation sector are indicated. The equations applied to estimate Ψest from CWSI were the following: y = −1.846x2 + 0.426x − 0.773 (stage II), y = −1.171x2 − 0.714 − 0.805 (stage III), and y = −1.558x − 1.144 (post-harvest).
Remotesensing 08 00039 g008

4. Discussion

To our knowledge, studies that published NWSB in heterogeneous crop surfaces are scarce. A linear NWSB was developed in this study for peach trees using all season data of two following years (Figure 2d). This NWSB was similar to those reported for other crops such as olive and sweet lime trees, but completely different than other crops (Table 4 [5,20,21,34,35,36,37]). Mandarine and orange trees were the only crops that reported a higher intercept of the NWSB than peach trees [21]. Therefore, it seems that well-watered peach trees displayed a lower transpiration rate than most other crops, leading an increase in Tc-Ta at a given VPD.
Table 4. Parameters of the NWSB (Tc-Ta = a·VPD + b) for different heterogeneous crops.
Table 4. Parameters of the NWSB (Tc-Ta = a·VPD + b) for different heterogeneous crops.
SpeciesCultivarabReference
PeachRoyal Glory−1.713.87This study
NectarineIndependence−1.690.68[34]
AppleRoyal Gala−3.901.00[35]
Sweet limeSwing−1.743.61[36]
PistachioKerman−1.332.44[20]
OliveArbequina−2.053.97[37]
MandarinClemenvilla−0.504.06[21]
OrangePowell−0.384.58[21]
GrapevinesChardonnay−1.392.16[5]
The relationship between Tc-Ta and VPD for well-watered peach trees presented a slightly different phenological response due to the higher intercept in the stage II regression (Figure 2). This implied that for a given increase in VPD, stage II had less transpirational cooling than stages III and post-harvest. Morphological and hydraulic consideration on trees can explain part of these seasonal differences. For instance, it is well-known that maximum transpiration rates occurs just before harvest, especially during the period of rapid fruit growth, and then drops off after the fruit has been removed [38]. Despite this, we did not detect different responses between stage III and post-harvest in the relationship Tc-Ta vs. VPD. Differences in water use may explain the dissimilarities between stages II and III. However, despite significance in the phenology, differences were not very large and did not have a negative effect on the estimation of CWSI. In fact, the relationships between CWSI and ΨL at different phenological stages exhibited the same response either taking into account different NWSB for each phenological stage or a common NWSB for the entire season (Figure 4). This means that CWSI for peach trees can be easily estimated throughout a growing season using the same lower and upper limits and without considering the phenological stage.
Figure 9. Relationship between stomatal conductance and (a) leaf water potential (ΨL) and; (b) CWSI in nectarines throughout the 2013 growing season. All relationships were significant (p < 0.0001).
Figure 9. Relationship between stomatal conductance and (a) leaf water potential (ΨL) and; (b) CWSI in nectarines throughout the 2013 growing season. All relationships were significant (p < 0.0001).
Remotesensing 08 00039 g009
A phenological effect was detected in the CWSI vs. ΨL relationships (Figure 4). These differences may depend on osmotic potential and leaf turgor seasonal changes. Marsal et al. [26] reported phenological differences for peach trees in osmotic leaf water potential at turgor loss point (Ψπ°), and osmotic potential at full turgor (Ψπ100). Both, Ψπ° and Ψπ100 trend to decrease as crops developed. Therefore, this hypothesis may explain the phenological differences detected in the relationship between CWSI and ΨL, where for a specific CWSI value, ΨL was more negative as the crop developed (Figure 4). A similar behavior was also reported in grapevines by Bellvert et al. [27]. In contrast to post-harvest, the pre-harvest CWSI vs. ΨL relationships were curvilinear. This indicates a lower ΨL sensitivity to CWSI in pre-harvest than in post-harvest stages. The reason for this curvature initiation is due in part to progressive stomatal closure in response to ΨL at early stages. In fact, there was a clear distinction between ΨL influences on stomatal conductance (gs) in the different phenological stages (Figure 9a). The relationship between ΨL and gs shows a progressive decrease in ΨL from a certain threshold of gs at pre-harvest stages (stages II and III), while it is linear at post-harvest. The same level of gs, corresponded to lower ΨL values at post-harvest than at pre-harvest stages. This indicated a higher control of water loss during pre-harvest stages than post-harvest, which seems consistent with the phenological sensitivity of gs to ΨL. ΨL at turgor loss from water (when gs approaches zero) approached levels of −3.0 and −2.9 MPa respectively for pre- and post-harvest stages. These values are in agreement with those reported by Marsal et al. [26] in peaches. The relationship between CWSI and gs also indicated a different phenological response (Figure 9b). A specific gs corresponded with lower CWSI values at pre-harvest than post-harvest stage. It has been reported that fruit removal significantly decreases the net photosynthesis rate, stomatal conductance and transpiration rate, and increased leaf temperature around midday [39]. Therefore, the higher CWSI values of post-harvest can be explained because of the reduction on the transpiration rate and thereby leaf evaporative cooling.

4.1. Validations with Nectarine and Saturn Peach

One goal of this research was to find out if the relationship between CWSI and ΨL had the same response among peach cultivars. This regression was also validated for nectarine in 2012 and 2013 and for Saturn peach in 2014. Figure 6 shows that data collected for nectarine in two years agreed with the seasonal CWSI-ΨL relationships of peach trees. Also, data of Saturn peaches followed the same pattern. Although a slight decrease in the R2 was detected when data for three cultivars was pooled in comparison to using only data for peach trees, the relationships were significant and R2 ranged from 0.52 (stage II) to 0.78 (stage III) (Figure 6). Therefore, it seems that contrary to grapevines, where differences between cultivars were detected [27], a common algorithm for each specific phenological stage can be used to estimate ΨL in the three peach cultivars.
The one-to-one relationships between estimated and observed ΨL were again plotted using the common algorithms for the three peach cultivars (Figure 7). Generally, despite these regressions being less well fitted to 1:1 than the peach tree data (Figure 5), differences were not significant. The corresponding regression of stage II had the poorest fitting (Figure 7a). It seems that ΨL values above −1.0 MPa were difficult to estimate. This is because the CWSI-ΨL relationship at stage II is curvilinear and there is little response of ΨL to changes of CWSI below the threshold of 0.3 (Figure 6). This might make it difficult to estimate ΨL for well-watered peaches during the early season. The RMSE between observed and estimated ΨL at different phenological stages ranged from 0.18 to 0.21 MPa. However, it seems that the RMSE had a different response depending on ΨL, where trend to decrease as ΨL was higher (Table 5). The overall relationship between averaged ΨL and RMSE for flights from all three years was significant with a R2 of 0.91.
Table 5. Relationship between the Root Mean Square Error (RMSE) between observed and estimated leaf water potential at different leaf water potential (ΨL) ranges for the three Prunus persica L. cultivars.
Table 5. Relationship between the Root Mean Square Error (RMSE) between observed and estimated leaf water potential at different leaf water potential (ΨL) ranges for the three Prunus persica L. cultivars.
ΨL Range (MPa)RMSE
−2.0 to −2.60.26
−1.6 to 1.90.25
−1.3 to −1.80.22
−1.0 to −1.20.18
−0.6 to −0.90.18

4.2. Minimum Pixel Size for Water Stress Detection

One of the limitations to successfully detect crop water status through thermal imagery is the pixel size. In grapevines, it was demonstrated that the minimum pixel size is at least 0.30 m [17]. However, canopy area and width of grapevines is much less than in stone fruit trees. Therefore, it was expected that peach and nectarine water status could be remotely detected at poorer spatial resolutions. The results indicated that, in peach and nectarine trees such as those from this study (tree crown area between 3.0 to 5.0 m2), it is necessary to obtain thermal images of at least 0.80 m pixel size (Table 3). However, it is also important to mention that the accuracy increases as the pixel size decreases.

4.3. Monitoring Irrigation Based on Crop Water Status Maps

This study demonstrated that CWSI is a suitable indicator for water status monitoring in peach and nectarine orchards in semi-arid conditions, and also for manage irrigation in precision agriculture. The acquisition of Ψest maps for an entire growing season allowed farmers to identify spatial differences in terms of water status and to carry out a more efficient differential irrigation management by either identifying sensitive areas or by setting different thresholds for irrigation. It has also been demonstrated that the adoption of RDI strategies can generate considerable water savings without having a negative impact on yield. Therefore, the use of remotely sensed maps of Ψest appears as a valuable tool for implementing RDI strategies in commercial orchards.

5. Conclusions

It was concluded that CWSI derived from remotely sensed thermal imagery is a suitable indicator for water stress monitoring in peach species. This study demonstrated that calculation of CWSI in Prunus Persica L. cultivars has not a different seasonal response and a common “non-water-stressed baseline” can be used throughout the growing season. However, for remote estimations of ΨL it is necessary to take into account the different seasonal CWSI-ΨL response. It is also important to underline that estimations of ΨL are robust for different growing seasons and that it has the same response in the three studied Prunus persica L. cultivars. Therefore, the applicability of these results are presented as a promising and powerful tool for assessing variability of crop water stress in heterogeneous crop orchards.

Acknowledgments

The funding of this research came from the Spanish Ministry of Economy and Competitiveness, under Project INNPACTO IPT-2011-1786-060000. We are grateful to Codorniu winery and Sorigué, S.A. for their support in this study. We thank RS Aviation, and in particular Mr Robert Pedra for his contribution to the acquisition of images. Technicians from the Efficient Use of Water program in IRTA and Quantalab (IAS-CSIC) are also acknowledged for their technical support in the field campaigns.

Author Contributions

All authors from IRTA and IAS-CSIC contributed to developing the overall research concept and data interpretations. Pablo J. Zarco-Tejada processed the image collections. Joaquim Bellvert collected and analyzed field data, as well as wrote the majority of the manuscript with contributions from all the remaining authors. Susan L. Ustin contributed in the final revisions of the manuscript.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Fereres, E.; Soriano, M.A. Deficit irrigation for reducing agricultural water use. J. Exp. Bot. 2007, 58, 147–159. [Google Scholar] [CrossRef] [PubMed]
  2. Behboudian, M.H.; Marsal, J.; Girona, J.; Lopez, G. Quality and yield responses of deciduous fruits to reduced irrigation. Hort. Rev. 2011, 38, 149–189. [Google Scholar]
  3. Lopez, G.; Behboudian, M.H.; Echeverria, G.; Girona, J.; Marsal, J. Instrumental and sensory evaluation of fruit quality for “Ryan’s sun” peach grown under deficit. Irrig. Hort. Technol. 2011, 21, 712–719. [Google Scholar]
  4. Sepulcre-Cantó, G.; Zarco-Tejada, P.J.; Jimenez-Muñoz, J.; Sobrino, J.; de Miguel, E.; Villalobos, F.J. Detection of water stress in a olive orchard with thermal remote sensing imagery. Agric. For. Meteorol. 2006, 136, 31–44. [Google Scholar]
  5. Bellvert, J.; Zarco-Tejada, P.J.; Marsal, J.; Girona, J.; González-Dugo, V.; Fereres, E. Vineyard irrigation scheduling based on airborne thermal imagery and water potential thresholds. Aust. J. Grape Wine Res. 2015. [Google Scholar] [CrossRef]
  6. Tanner, C.B. Plant temperatures. Agron. J. 1963, 55, 210–211. [Google Scholar] [CrossRef]
  7. Fuchs, M.; Tanner, C.B. Infrared thermometry of vegetation. Agron. J. 1966, 58, 297–601. [Google Scholar] [CrossRef]
  8. Idso, S.B.; Jackson, R.D.; Reginato, R.J. Extending the degree day concept of plant phenological development to include water stress effects. Ecology 1978, 59, 431–433. [Google Scholar] [CrossRef]
  9. Idso, S.B.; Jackson, R.D.; Pinter, P.J.; Reginato, R.J.; Hatfield, J.L. Normalizing the stress-degree day parameter for environmental variability. Agric. Meteorol. 1981, 24, 45–55. [Google Scholar] [CrossRef]
  10. Jackson, R.; Reginato, R.; Idso, S.B. Wheat canopy temperature: A practical tool for evaluating water requirements. Water Resour. Res. 1977, 13, 651–656. [Google Scholar] [CrossRef]
  11. Jackson, R.; Idso, S.; Reginato, R.; Pinter, P.J. Canopy temperature as a crop water stress indicator. Water Resour. Res. 1981, 17, 1133–1138. [Google Scholar] [CrossRef]
  12. Jackson, R. Canopy temperature and crop water stress. Adv. Irrig. 1982, 1, 43–85. [Google Scholar]
  13. Wanjura, D.F.; Hatfield, J.L.; Upchurch, D.R. Crop water stress index relation-ships with crop productivity. Irriig. Sci. 1990, 11, 93–99. [Google Scholar]
  14. Irmak, S.; Haman, D.Z.; Bastug, R. Determination of crop water stress index for irrigation timing and yield estimation of corn. Agron. J. 2000, 92, 1221–1227. [Google Scholar] [CrossRef]
  15. Kar, G.; Kumar, A. Surface energy fluxes and crop water stress index in ground-nut under irrigated ecosystem. Agric. For. Meteorol. 2007, 146, 94–106. [Google Scholar] [CrossRef]
  16. Möller, M.; Alchanatis, V.; Cohen, Y.; Meron, M.; Tsipris, J.; Naor, A.; Ostrovsky, V.; Sprintsin, M.; Cohen, S. Use of thermal and visible imagery for estimating crop water status of irrigated grapevine. J. Exp. Bot. 2007, 58, 827–838. [Google Scholar] [CrossRef] [PubMed]
  17. Bellvert, J.; Zarco-Tejada, P.J.; Girona, J.; Fereres, E. Mapping crop water stress index in a “Pinot-noir” vineyard: Comparing ground measurements with thermal remote sensing imagery from an unmanned aerial vehicle. Precis. Agric. 2014, 15, 361–376. [Google Scholar] [CrossRef]
  18. Ben-Gal, A.; Agam, N.; Alchanatis, V.; Cohen, Y.; Yermiyahu, U.; Zipori, I.; Presnov, E.; Sprintsin, M.; Dag, A. Evaluating water stress in irrigated olives: Correlation of soil water status, tree water status, and thermal imagery. Irrig. Sci. 2009, 27, 367–376. [Google Scholar] [CrossRef]
  19. Berni, J.J.; Zarco-Tejada, P.J.; Sepulcre-Cantó, G.; Fereres, E.; Villalobos, F. Mapping canopy conductance and CWSI in olive orchards using high resolution thermal remote sensing imagery. Remote Sens. Environ. 2009, 113, 2380–2388. [Google Scholar] [CrossRef]
  20. Testi, L.; Goldhamer, D.A.; Iniesta, F.; Salinas, M. Crop water stress index is a sensitive water stress indicator in pistachio trees. Irrig. Sci. 2008, 26, 395–405. [Google Scholar] [CrossRef]
  21. Gonzalez-Dugo, V.; Zarco-Tejada, P.J.; Fereres, E. Applicability and limitations of using the crop water stress index as an indicator of water deficits in citrus orchards. Agric. For. Meteorol. 2014, 198–199, 94–104. [Google Scholar] [CrossRef]
  22. Wang, D.; Gartung, J. Infrared canopy temperature of early-ripening peach trees under postharvest deficit irrigation. Agric. Water Manag. 2010, 97, 1787–1794. [Google Scholar] [CrossRef]
  23. Cohen, Y.; Alchanatis, V.; Meron, M.; Saranga, Y.; Tsipis, J. Estimation of leaf water potential by thermal imagery and spatial analysis. J. Exp. Bot. 2005, 417, 1843–1852. [Google Scholar] [CrossRef] [PubMed]
  24. Gonzalez-Dugo, V.; Zarco-Tejada, P.J.; Nicolás, E.; Nortes, P.A.; Alarcón, J.J.; Intrigliolo, D.S.; Fereres, E. Using high resolution UAV thermal imagery to assess the variability in the water status of five fruit tree species within a commercial orchard. Precis. Agric. 2013, 14, 660–678. [Google Scholar] [CrossRef]
  25. McCutchan, H.; Shackel, K. Stem-water potential as a sensitive indicator of water stress in prune trees (Prunus domestica L. cv. French). J. Am. Soc. Hortic. Sci. 1992, 117, 607–611. [Google Scholar]
  26. Marsal, J.; Girona, J. Relationship between leaf water potential and gas exchange activity at diferent phenological stages and fruit loads in peach trees. J. Am. Soc. Hortic. Sci. 1997, 122, 415–421. [Google Scholar]
  27. Bellvert, J.; Zarco-Tejada, P.J.; Girona, J.; Marsal, J.; Fereres, E. Seasonal evolution of crop water stress index in grapevine varieties determined with high resolution remote sensing thermal imagery. Irrig. Sci. 2015, 33, 81–93. [Google Scholar] [CrossRef]
  28. Allen, R.G.; Pereira, L.S.; Raes, D.; Smith, M. Crop Evapotranspiration. Guidelines for Computing Crop Water Requirements; FAO Irrigation and Drainage Paper No. 56; FAO: Rome, Italy, 1998. [Google Scholar]
  29. Doorenbos, J.; Pruitt, W. Guidelines for Predicting Crop Water Requirements; FAO Irrigation and Drainage Paper No. 24; FAO: Rome, Italy, 1977; p. 144. [Google Scholar]
  30. Berni, J.A.; Zarco-Tejada, P.J.; Suarez, L.; Fereres, E. Thermal and narrow-band multispectral remote sensing for vegetation monitoring from an unmanned aerial vehicle. IEEE Trans. Geosci. Remote Sens. 2009, 47, 722–738. [Google Scholar] [CrossRef]
  31. Zarco-Tejada, P.J.; González-Dugo, V.; Berni, J.A.J. Fluorescence, temperature and narrow-band indices acquired from a UAV platform for water stress detection using a micro-hyperspectral imager and a thermal camera. Remote Sens. Environ. 2012, 117, 322–337. [Google Scholar] [CrossRef]
  32. Turner, N.C.; Long, M.J. Errors arising from rapid water loss in the measurement of leaf water potential by pressure chamber technique. Aust. J. Plant Physiol. 1980, 7, 527–537. [Google Scholar] [CrossRef]
  33. SAS. Enterprise Guide Version 4.2; SAS Institute Inc.: Cary, NC, USA, 2002. [Google Scholar]
  34. Tormann, H. Canopy temperature as a plant water stress indicator for nectarines. S. Afr. J. Plant Soil 1986, 3, 110–114. [Google Scholar] [CrossRef]
  35. Andrews, P.K.; Chalmers, D.J.; Moremong, M. Canopy-Air temperature differences and soil water as predictors of water stress of apple trees grown in a humid, temperate climate. J. Am. Soc. Hortic. Sci. 1992, 117, 453–458. [Google Scholar]
  36. Sepaskhah, A.R.; Kashefipour, S.M. Relationships between leaf water potential, CWSI, yield and fruit quality of sweet lime under drip irrigation. Agric. Water Manag. 1994, 25, 13–22. [Google Scholar] [CrossRef]
  37. Bellvert, J.; Zarco-Tejada, P.J.; Girona, J.; Gonzalez-Dugo, V.; Fereres, E. A tool for detecting crop water status using airborne high-resolution thermal imagery. In Sustainable Irrigation and Drainage V: Management, Technologies and Policies; WIT Transactions on Ecology and The Environment: Poznan, Poland, 2014; pp. 25–31. [Google Scholar]
  38. Kramer, P.J. Water Relations of Plants; Academic Press: New York, NY, USA, 1983. [Google Scholar]
  39. Li, W.D.; Li, S.H.; Yang, S.H.; Yang, J.M.; Zheng, X.B.; Li, X.D.; Yao, H.M. Photosynthesis in response to sink-source manipulations during different phenological stages of fruit development in peach trees: Regulation by stomatal aperture and leaf temperature. J. Hortic. Sci. Biotechnol. 2005, 80, 481–487. [Google Scholar]

Share and Cite

MDPI and ACS Style

Bellvert, J.; Marsal, J.; Girona, J.; Gonzalez-Dugo, V.; Fereres, E.; Ustin, S.L.; Zarco-Tejada, P.J. Airborne Thermal Imagery to Detect the Seasonal Evolution of Crop Water Status in Peach, Nectarine and Saturn Peach Orchards. Remote Sens. 2016, 8, 39. https://doi.org/10.3390/rs8010039

AMA Style

Bellvert J, Marsal J, Girona J, Gonzalez-Dugo V, Fereres E, Ustin SL, Zarco-Tejada PJ. Airborne Thermal Imagery to Detect the Seasonal Evolution of Crop Water Status in Peach, Nectarine and Saturn Peach Orchards. Remote Sensing. 2016; 8(1):39. https://doi.org/10.3390/rs8010039

Chicago/Turabian Style

Bellvert, Joaquim, Jordi Marsal, Joan Girona, Victoria Gonzalez-Dugo, Elías Fereres, Susan L. Ustin, and Pablo J. Zarco-Tejada. 2016. "Airborne Thermal Imagery to Detect the Seasonal Evolution of Crop Water Status in Peach, Nectarine and Saturn Peach Orchards" Remote Sensing 8, no. 1: 39. https://doi.org/10.3390/rs8010039

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop