Introduction

Orchards and vineyards are increasingly irrigated. Knowing their water requirements is essential to estimate their irrigation requirements, planning and management of crop water use, assessing water resources availability and demand at basin level, as well as developing hydrologic studies. Accuracy in evapotranspiration (ET) estimates is necessary, mainly when water scarcity prevails, and because sustainable irrigation requires not exceeding crop demand to break the trend for water over-use (Pereira et al. 2009; Wada and Bierkens 2014; Müller Schmied et al. 2016). In addition, related challenges are becoming more difficult due to a continuously increased demand for food to nourish an ever-growing population, increasing drought occurrences, and climate change. As recently reviewed by Pereira (2017), high water use performance and productivity, as well as water conservation and saving in irrigation, require solutions that need improved knowledge of crop evapotranspiration and water use.

The application of water conservation and saving, in addition to the knowledge of water needs and their relations to growth and yield, also require institutional interventions, sectoral policies and new technologies that support improved irrigation management and performance by farmers and sustainable, eco-friendly use of water for food production (Pereira et al. 2009). Literature on management of fruit trees and grapevines is quite extensive, namely relative to water management, and particularly aimed at defining deficit irrigation (DI) strategies.

The concept of standard crop coefficient implies its determination in the absence of water stress, or other stress conditions. However, research on tree and vine crops is demonstrating that the best crop management does not correspond to the full satisfaction of crop water demand, but to the adoption of controlled water deficit at given periods, or in selected modes during the crop cycle aiming that yields are less affected (Chaves et al. 2010; Rallo et al. 2017; Romero et al. 2022) and quality is improved (e.g., López-Urrea et al. 2012). The concept of eustress may better describe such conditions than deficit irrigation (Paço et al. 2019; Rallo et al. 2021). Expanding basic and accurate information on crop water needs is paramount to improve water use and irrigation management, particularly of fruit trees and vines.

Crop evapotranspiration (ETc) is typically computed or modeled using the well-known FAO56 calculation procedure (Allen et al. 1998), which uses the simple Kc–ETo approach to compute ETc, or alternatively Kcb–ETo to compute crop transpiration (Tc), i.e., the product of a crop coefficient (Kc) by the grass reference evapotranspiration (ETo), or the product of a basal crop coefficient (Kcb) by ETo. The latter represents the actual evaporative demand of the atmosphere, while Kc (ratio ETc/ETo) represents an integration of the effects of the main characteristics that distinguish, in terms of the energy balance, the grass reference crop from the crop under study (Allen et al. 1998; Pereira et al. 1999). Adopting the Kc–ETo approach is simple but requires the application of accurate measurements and computations, particularly when deriving Kc values for a crop using field observations (Allen et al. 2011; Pereira et al. 2021a, b).

Standard, transferable crop coefficients must be obtained from accurate ETc field measurements under non- or eustress conditions, and ETo computed with the FAO-PM ETo (Allen et al. 1998), or the ASCE-PM ETo (Allen et al. 2005). Other equations whose results relate well to those of the FAO-PM ETo equations may also be used. Adopting fixed grass parameters for aerodynamic and surface resistance in FAO-PM ETo equation provides for the crop coefficients to be crop specific parameters that express consistently the relation between the aerodynamic and surface resistances of the considered crop with those of the grass reference crop (Pereira et al. 1999). This is particularly challenging for vines and fruit trees due to their canopy architecture and incomplete ground cover.

Accurate standard, transferable, and updated Kc values obtained from the current literature review require that related ETc data collection, models and related model calibrations, as well as experimental set-ups were exempt of biases caused by experimental flaws (Allen et al. 2011). Following the methodology adopted in studies focused on vegetable and field crops (Pereira et al. 2021a, b), the selected references were checked to ensure that sufficient descriptions of ETc measurement practices, crop management, and related production environment were provided. They were also checked to detect possible computational flaws and shortcomings in data handling, as well as in model calibration and validation. In addition, the possible influence of advection was considered (e.g., Wang et al. 2019) since related Kc∕Kcb values are then of local value only, thus not transferable. Nonetheless, for several crops, the collected information was scarce.

Few studies reports on tabulated standard Kc∕Kcb of trees and vine crops. The first is FAO56 guidelines (Allen et al. 1998), whose Kc∕Kcb values continue to be the main reference for trees and vine crops. Later, Allen and Pereira (2009) suggested the A&P approach to determine Kc∕Kcb from the fraction of ground cover and height and tabulated the related values. Jensen and Allen (2016) tabulated again Kc∕Kcb for woody perennials. The A&P approach was tested for more crops and the resulting Kc∕Kcb were reported to support further use of the A&P approach (Pereira et al. 2020b, 2021c). Finally, Kc∕Kcb updated values were tabulated by Rallo et al. (2021). Excellent Kcb and Kc results were predicted from the field observed fraction of ground cover and height (Allen and Pereira 2009; Pereira et al. 2020b, 2021c). The A&P approach is particularly interesting for woody and incomplete cover crops, e.g., fruit trees and vines.

The Kc–ETo method is the most common in practice but not in research. The selected literature reports numerous applications of the Kc–ETo method using the soil–water balance (SWB) based on a variety of soil water content (SWC) sensors, which accuracy was reviewed by Evett et al. (2012a) and computation procedures, including a diversity of calibrated models, were recently reviewed (Pereira et al. 2020a). Diverse field measuring approaches of actual ETc or Tc (ETc act and Tc act) are reported such as weighing, drainage, and water-table lysimeters (WL, DL, and WTL), as reviewed by Allen et al. (1991) and Evett et al. (2012c, 2016), the eddy covariance systems (EC, Cammalleri et al. 2013a), the Bowen ratio energy balance systems (BREB, Hu et al. 2014), sap-flow (SF, Fernández et al. 2008), and remote sensing vegetation indices (RS-VI, Pôças et al. 2020). Allen et al. (2011) and Evett et al. (2012a) reviewed these methods for accuracy in ETc act measurements, and various authors also compared diverse methods for accuracy (Sánchez et al. 2019). Methods not requiring Kc nor ETo are also often used such as remote sensing surface energy balance algorithms (RS-SEB, Karimi and Bastiaanssen 2015), the Penman–Monteith combination equation (PM-eq, Monteith 1965), that uses actual aerodynamic and canopy resistances, the two-source Shuttleworth–Wallace method (SW, Shuttleworth and Wallace (1985), that also relies on those resistances, or the Priestley–Taylor equation (PT, Priestley and Taylor (1972), which uses specific coefficients different of Kc. These methods may provide for Kc when ETo is reported in addition to ETc act.

Recent advances in sensors, communications and information technologies did allow the implementation of tools to support irrigation and water management decisions, e.g., the “internet of things” (García et al. 2020; Raj et al. 2021; Abu et al. 2022). Tools focusing on irrigation may refer to water saving practices (Pereira et al. 2009; Jovanovic et al. 2020), to modeling growth and yield (Villalobos et al. 2006; Rahmati et al. 2018) or, most often, to SWB models (Pereira et al. 2020a). However, only few report computer software models (e.g., Rosa et al. 2012a, b; Šimůnek et al. 2016). The use of all those tools to support irrigation management requires precise knowledge of crop water requirements. The various methods need to be known for their accuracy requirements while Tables summarizing the information provided in literature need to be completed with indication of field methods used to derive Kc/Kcb.

The objective of this paper, in line with the previous review and addressing particularly Mediterranean and warm temperate fruit and leaf tree and vine crops, consists of reviewing updated single and basal crop coefficient values (Kc and Kcb) obtained under near-pristine eustress conditions and use the available Kc and Kcb data for tabulating standard, transferable Kc and Kcb values. The current review is expected to identify the main results of recent research on standard Kc and Kcb values, assessing the methodologies then used and their range of variation. The selected base data and collected values are, therefore, summarized and tabulated to support readers’ information on tabulated Kc and Kcb values. Section “Materials and methods” focus on requirements for accuracy of the ET methods reported on transferability requirements of standard Kc and Kcb values, and on the methodologies used to select and tabulate the standard crop coefficient values. Section “Table and wine grapes” and following consist of presenting and discussing the literature review relative to the derivation of Kc act and Kcb act of wine and table grapes, olive, citrus, avocado, loquat, persimmon, and tea, followed by the tabulation of the respective standard Kc and Kcb values. Conclusions and recommendations consist of the last Section.

Materials and methods

The FAO56 evapotranspiration method. Crop coefficients and requirements for transferability

Adopting the FAO56 method, crop evapotranspiration, ETc (mm d−1) is estimated by multiplying the grass reference ETo (mm d−1), by a crop coefficient, Kc (dimensionless):

$${\text{ET}}_{{\text{c}}} = \,{\text{ET}}_{{\text{o}}} {\text{K}}_{{\text{c}}}$$
(1)

ETo is defined as the evapotranspiration of a grass reference crop which is a hypothetical crop with height of 0.12 m, a surface resistance of 70 s m−1, and an albedo of 0.23, closely resembling an extensive surface of green grass of uniform height, actively growing, adequately watered, and well covering the ground (Allen et al. 1998). The daily ETo is computed with the PM-ETo equation (Eq. 2), obtained by parameterizing the Penman–Monteith combination equation for that grass crop with fixed and well-defined aerodynamic and surface resistance terms (Allen et al. 1998; Pereira et al. 1999). Daily grass reference evapotranspiration is then obtained with the following equation:

$${\mathrm{ET}}_{o}= \frac{0.408\Delta \left({\mathrm{R}}_{\mathrm{n}}-\mathrm{G}\right)+\upgamma \frac{900}{\mathrm{T}+273} {\mathrm{u}}_{2} \left({\mathrm{e}}_{\mathrm{s}}- {\mathrm{e}}_{\mathrm{a}}\right) }{\Delta +\upgamma \left(1+0.34 {\mathrm{u}}_{2}\right)},$$
(2)

where Δ is the slope of the saturation vapor pressure–temperature curve at mean air temperature (kPa °C−1), (Rn–G) is the available energy at the vegetated surface (MJ m−2 d−1), γ is the psychrometric constant (kPa °C−1), T is mean daily air temperature (°C), u2 is mean daily wind speed (m s−1) at 2 m height and (es−ea) is the vapor pressure deficit (VPD) of the atmosphere (kPa). The PM-ETo equation considers only vertical fluxes of heat and vapor. Thus, ETo incorporates most of the weather and related energy effects and then represents the evaporative demand of the atmosphere. Since Kc is the ratio between ETc and ETo (Eq. 1), its variations should mainly be attributed to the specific crop characteristics and only for a limited extent to the climate, which enables the transfer of standard Kc values between locations and climates when local and/or regional advection is excluded.

Apart from the FAO-PM-ETo equation, other alternative equations have been tested to calculate ETo, either with full or limited weather data sets. Processes with full data sets have the tendency to overlook the conceptual framework (Pereira et al. 2015). For reduced data sets, the Hargreaves–Samani equation (Hargreaves and Samani 1985) and the FAO PM temperature (FAO-PMT) method have been widely used; consolidated methodologies are discussed and described by (Paredes et al. 2020), as well as the use of reanalysis weather data and of geostationary satellite products (Paredes et al. 2021). However, the use of alternative approaches requires the scrutiny of input data and ETo results since processes are not linear. Therefore, for scientific research studies intending to derive standard transferable crop coefficients, the FAO-PM-ETo Eq. (2) should be used.

The crop coefficient represents an integration of the effects of three primary characteristics that distinguish any crop from the reference one: crop height, that affects roughness and aerodynamic resistance (ra); bulk crop–soil surface resistance (rs), which relates to leaf area, the fraction of ground covered by the vegetation (fc), leaf age and condition, degree of stomatal control, and soil surface wetness; and albedo of the crop–soil surface influencing the net radiation, that is determined by the fraction of ground covered by vegetation, and soil surface wetness (Allen et al. 1998).

Two Kc approaches are considered (Allen et al. 1998): one consists of a time-averaged single Kc, which includes multi-day effects of evaporation from the soil in addition to plant transpiration, whereas the second refers to the dual Kc, sum of the basal crop coefficient (Kcb) and the soil evaporation coefficient (Ke). These coefficients represent, respectively, the ratios of crop transpiration (Tc), or soil evaporation (Es), to ETo. Therefore, Kc = Kcb + Ke with Kcb = Tc/ETo and Ke = Es/ETo.

Various authors have developed models or procedures for partitioning ET into Tc and Es. However, the FAO56 approach (Allen et al. 1998, 2005) has been successfully used and implemented in various SWB models such as SIMDualKc (Rosa et al. 2012a), whose applications to vineyards (Fandiño et al. 2012; Silva et al. 2021; Darouich et al. 2022b), olive (Paço et al. 2014, 2019; Puig-Sirera et al. 2021; Ramos et al. 2023), and citrus (Rosa 2018; Peddinti and Kambhammettu 2019; Darouich et al. 2022a; Ramos et al. 2023) are reported herein.

For transferability purposes, FAO56 adopted the concept of standard Kc and potential ETc (Allen et al. 1998; Pereira et al. 2015), which refer to well-watered and pristine/eustress cropping conditions and are distinct of actual field conditions, often under-optimal due to insufficient (or non-uniform) irrigation, crop density, salinity, agronomic practices and soil management. The tabulated Kc, therefore, must refer exclusively to the standard Kc. For tree and vine crops, the standard Kc refers to adopting crop-specific eustress practices, i.e., limited stress practices that do not, or minimally, impose reduction of the maximal yield. Under water and salt stress conditions, ETc gives place to the actual crop ET (ETc act), with Kc replaced by the actual Kc act or, using the dual approach, by Ks Kcb + Ke:

$${\text{ET}}_{{\text{c act}}} = {\text{ K}}_{{\text{s}}} \,{\text{ET}}_{{\text{c}}} = {\text{ K}}_{{\text{s}}} \,{\text{K}}_{{\text{c}}} \,{\text{ET}}_{{\text{o}}} = \, \left( {{\text{K}}_{{\text{s}}} \,{\text{K}}_{{{\text{cb}}}} + {\text{ K}}_{{\text{e}}} } \right){\text{ ET}}_{{\text{o}}} ,$$
(3)

where Ks (0–1.0) is the stress coefficient. Ks depends upon the sufficiency of available soil water to maintain the crop ET rate, i.e., Ks = 1.0 for pristine conditions for maximal yield. This concept eases a consistent estimation and transferability of measured standard Kc and avoids the need to define multiple Kc values for the same crop depending upon the various water and crop management practices adopted by the growers that cause Ks < 1.0 and Kc act values to vary widely, contrarily to the standard Kc. Plot level use of crop coefficient-based simulations can be backed up by soil and plant water status measurements to detect water stress conditions (e.g., leaf or stem water potential) and to support the use of models.

The estimation of Kc act, assuming any value up to the standard Kc, may be performed using the A&P approach (Allen and Pereira 2009; Pereira et al. 2020b, 2021c). Kcb act is then computed from the fraction of ground cover and crop height (fc and h) while Ke is computed from the wetted fraction of exposed soil, 1−fc (FAO56, Allen et al. 1998). The A&P approach shall be used with observed fc and h and the parameters proposed in Pereira et al. (2020b, 2021c). It is advisable to compare Kc act = Kcb act+ + Ke with the standard Kc for computation control. Field and remote sensing methods for measuring fc and h are referred by those authors.

Evapotranspiration relies on the amount of energy available at the surface, resulting from the energy balance of that surface:

$$\lambda {\text{ET}} = \left( {{\text{R}}_{{\text{n}}} {-}{\text{G}}} \right){-}{\text{H}},$$
(4)

where λET is latent heat flux, or the energy available to the evaporation process, Rn is net radiation at the crop surface, G is soil heat flux density and H is sensible heat flux, with all terms expressed in MJ m−2 day−1. The energy balance imposes physical limits to the evaporation process resulting that the upper limits to Kc are approximately 1.2 in sub-humid regions and 1.2–1.4 in arid regions (Allen et al. 2011). Higher values might result from errors in ET measurement, in weather data for ETo calculation, in data processing procedure, or may be due to advective energy. Awareness of such upper limits of Kc is extremely important; conditions where measurements were acquired or those from where Kc’s are meant to be applied must be considered, namely in terms of advection: If the λET term of the surface energy balance equation (Eq. 4) results in a value higher than Rn–G, the surface is receiving sensible heat downwards, instead of just losing it by convection to the atmosphere. Therefore, a larger amount of energy will be available for the process of evapotranspiration. However, there is an upper boundary to ETc, imposed by limitations in aerodynamic transfer and equilibrium forces over a vegetated surface (Allen et al. 2011). Then, limits apply and, in general, Kc ≤ 1.2 except in the presence of advection. Advection conditions can limit transferability of crop coefficients, either because they were determined under advection conditions or they are to be applied in such conditions.

Advection can result from the small dimension of the stand under consideration, not providing adequate conditions for the development of a boundary layer in equilibrium with the surface, or by favoring a “clothesline effect”, where stand vegetation is more exposed to atmosphere drive than the surrounding vegetation (Allen et al. 2011). Advection can also result from inadequate field measurement conditions, e.g., when lysimeters are not correctly set, causing local and micro-scale advection, or a “clothesline effect”; or when fetch conditions in EC and BREB systems are not observed, or data quality selection criteria against wind direction/fetch are not applied (Hu et al. 2014). Under advection, H decreases to very small values, given the downward advective H flux and, therefore, λET ≥ (Rn−G). Hence, it is expected that under advection conditions, and over small stands of vegetation, ETc would reach a much larger value (Allen et al. 2011), which is not the case for large stands, where limits for Kc near 1.2 apply.

Advection effects on ETc of woody crops are rarely reported in literature. However, since trees and vines do not attain full crop development due to pruning and training and are partial cover crops, in the absence of advection, Kc values should not surpass 1.2 (Rallo et al. 2021), but under advective conditions much larger transpiration and larger soil evaporation values may be observed (Kool et al. 2018; Wang et al. 2019); nevertheless, too much large Kc values are reported in literature without signaling the occurrence of advection. For application in small and isolated areas of vegetation, Kc can exceed the limits for grass reference (1.2–1.4), while for large areas, or small areas surrounded by vegetation with similar roughness and soil water status, Kc values must stick to values equal or smaller than those limits (Allen et al. 2011).

The concepts of standard Kc and potential crop ET and related terminology are progressively being accepted by the user communities (Pereira et al. 2015). However, the standard Kc and Kcb values for tree and vine crops vary with the fraction of ground cover and height (Allen and Pereira 2009; Jensen and Allen 2016) due to crop age and crop management, particularly crop training. The present review has shown that satisfactorily accurate reported Kc and Kcb values for the same crop show dissimilarity among locations, which is due to differences in cultivar and rootstock, plant density, orchard management and pruning, training, as well as soil properties, irrigation method and strategy, soil–crop management practices and (Minacapilli et al. 2009; Cammalleri et al. 2013a; Marsal et al. 2014; Rallo et al. 2021). It is, however, possible to derive local, actual crop coefficients from fc and h of tree and vine crops (Pereira et al. 2020b, 2021c) when appropriate parameters are used. Kc variability due to weather is less important than causes referred above. Since most papers did not provide weather data on the experiment, the correction of Kc values for climate as proposed in FAO56 was not applied to literature reported Kc.

Accuracy of ET estimation and transferability of derived standard Kc and Kcb

Literature reporting field derived crop coefficients has shown diverse objectives and used quite different methodologies with variable accuracy, often aiming to just obtain Kc values for local use, which are not transferable. Results are frequently published without sufficient information relative to the methods and instrumentation used, or about the crop itself, the cropping practices and training, which causes difficulties to transferability. When the published material shown serious limitations to transferability, it was not used. Main limitations refer to:

  1. (1)

    Adopting other than the standard FAO or ASCE PM-ETo equation. Because Kc is defined as the ratio ETc/ETo, if ETo equation changes Kc also changes and the resulting Kc act is not usable to derive a standard Kc.

  2. (2)

    Using a Kc curve different from the standard segmented FAO Kc curve. Using a curve as a function of time, or a function of LAI, or else, there is no clear definition of the Kc (and Kcb) values for the initial, mid-season and end-season stages, respectively, Kc ini, Kc mid, and Kc end. Then, only approximate estimations of Kc ini, Kc mid, and Kc end could be made from the reported graphical data or, often more difficult, from tabulated information.

  3. (3)

    Using non-standard cultivation conditions. In case of using mulch for controlling Es, or active ground cover for fighting erosion result management-specific Kc values. When reported Kc values were insufficient to properly recognize the standard Kc values, papers could not be used.

  4. (4)

    Adopting deficit irrigation practices. Then, deviating from the desirable eustress conditions, the reported Kc act had only local value.

  5. (5)

    Reporting insufficient data and information on the experiment. Then, it was not possible to assume that field survey practices were adequate, or that data handling were properly performed. To avoid using poor data, papers were discarded or used with much caution.

  6. (6)

    Using Kc values transferred from other studies. Without field testing, papers were discarded, except a few review papers.

Field data acquisition processes must respect well-defined requirements. Field data sets used to derive Kc or Kcb were obtained using various field techniques. The requirements for data quality acquisition by these methods are extensively described in Allen et al. (2011), reviewed by Pereira et al. (2021a, b) and Rallo et al. (2021), and summarized hereafter.

Techniques that recur to soil water balance methods calculate ETc as the remaining term as commented (Evett et al. 2012a, 2012b; Pereira et al. 2020a). The main sources of error arise from the quantification of deep percolation and/or capillary rise. Other difficulties may arise from the different patterns of soil water extraction by the roots, namely for heterogeneous stands, as sampling processes may not represent adequately the stand. Techniques must consider: (i) a comprehensive characterization of soil hydraulic properties, (ii) representativeness of data in spatial and temporal terms, (iii) appropriate sensors calibration, (iv) uniform spatial wetting by irrigation, (v) consideration of deep percolation and capillary rise, (vi) root water extraction patterns, and (vii) sampling criteria. Accuracy of computation procedures depends upon the calibration of parameters and the adequacy of selected algorithms (Pereira et al. 2020a).

Weighing, drainage, and water-table lysimeters (WL, DL, and WTL) are often used for Kc derivation but their accuracy depends upon various issues (Allen et al. 1991; López-Urrea et al. 2006; Evett et al. 2016). Causes of inaccuracy include: (i) differences in cropping conditions inside and outside of the lysimeter relative to vigor and growth of vegetation; (ii) poor setting of the lysimeter, with dissimilar surrounding vegetation causing local advection or clothesline effects; (iii) insufficient fetch to establish the equilibrium boundary layer of air (EBL); (iv) lack of consideration of the area effectively used by the crop for ET, which may often occur with trees and vines; (v) large rim favoring heat transfer into the lysimeter.

The BREB method relies on the surface energy balance equation (Eq. 4) and requires measurements of air temperature and vapor pressure gradients at an appropriate level above the evaporating surface (Hu et al. 2014). The accuracy of the method relies strongly on representativeness of Rn and G measurements and on an adequate fetch for the establishment of the EBL. Main requirements for BREB data quality include: (i) large enough fetch; (ii) adequate positioning of sensors above the canopy to avoid the roughness sublayer; (iii) multiple Rn and G measurement points for heterogeneous or sparse crops. The EC method implies the knowledge of vertical wind speed and fluctuations around the mean of air temperature and humidity in vertical fluxes of sensible and latent heat, sampling statistically turbulent eddies (Cammalleri et al. 2013a). For accuracy, requirements include: (i) large enough fetch and adequate elevation of sensors; (ii) application of the required corrections; (iii) recognition of advection situations and taking of corrective actions, and (iv) correcting data for lack of closure of the energy balance equation, when needed.

The transpiration component in ETc is generally obtained by sap flow measurement systems that use heat as a tracer to measure the flux in the xylem of plants. These methods generally follow well the transpiration dynamics but require calibration for accurate results (Fernández et al. 2008; Siqueira et al. 2020). Sap flow measurements require: (i) a sensor calibration at each new application, (ii) because measurements are plant-based processes, scaling from plant to stand level is required, then also dealing with measurement representativeness, and (iii) an accurate estimate of conductive xylem area.

Remote sensing is increasingly used to calculate ETc, namely using surface energy balance models (Pôças et al. 2014, 2020; Karimi and Bastiaanssen 2015; Sánchez et al. 2019), currently largely used for Kc and ET calculation. Vegetation indices derived from satellite information or using UAV, require ground data for validation and are related to actual crop coefficients (Garrido-Rubio et al. 2020; Pôças et al. 2020). Inaccuracies in measuring crop ET and in computing ETo often result in high Kc values, commonly indicating that the corresponding energy use would largely exceed the energy available at the surface for evaporation (Allen et al. 2011) as referred early.

Methods adopted to select the papers

The review focused on articles published after the FAO56 guidelines (Allen et al. 1998), until March 2023. The search first targeted the articles that quoted FAO56 or that referred crop coefficients. Several search engines were used (e.g., Schooler google, Elsevier, Springer, Wiley, Csiro publishing, Scielo, Scopus) as well as different combination of keywords (crop coefficients, orchards, and names and scientific names of plants). Various languages were used for the search (English, Portuguese, Spanish, French, Italian and German). Because Insufficiencies and inaccuracies referred in the previous sections limit the transferability of reported Kc values, to update the tabulated Kc, it was necessary to operate a careful literature selection. Limitations relative to accuracy of data acquisition, the Kc curves or crop conditions obliged a careful review of published material as referred before, to check when the proposed Kc or Kcb were limited to local interest and/or represented non-standard experimental conditions, thus contrasting to Kc resulting from near-pristine eustress cropping practices. Thus, studies were selected when:

  • Adopted the FAO-PM-ETo equation (Allen et al. 1998) or the ASCE-ETo equation (Allen et al. 2005) or other ETo equations if their ratios to FAO-PM-ETo could be approximated.

  • Presented data referred to two or more experimental seasons, or studies having various treatments, so that it was possible to understand if the results were or not occasional. However, a few cases referring only one season were considered for Neglected and Underutilized Species (NUS) crops.

  • Descriptions of experiments are sufficient to consider their accuracy and that crops were at conditions close to non-stress or just eustress.

  • Adopted the FAO Kc curve, or a Kc–time curve that allowed to identify Kc or Kcb for the mid-season and, preferably, also for the initial and end season.

  • Papers describing field studies using BREB or EC systems, which reported upon the upwind fetch conditions and the energy balance closure.

  • Studies using SWB methods describing all the terms of the balance, not just focusing the upper soil depth, and providing for an adequate description of sensors used and location, frequency of observations, and the model calibration and validation, were selected.

  • Studies using lysimeters were accepted when adequate setting and management were referred, namely avoiding “oasis” and “cloth-line” effects, and correction of the evaporative surface when the tree/vine canopy exceeded the lysimeter surface (“bloom effect”).

  • Studies using remote sensing were considered when adequate ground observations for model or vegetation index calibration/validation was taken into consideration.

  • Studies reporting acceptable Kc values (Kc up to 1.30 and Kcb < Kc) unless convincing explanations were given.

The assumed criteria made it possible to select a good number of papers, covering numerous species, developed in a variety of countries and regions, and in quite reasonable quality conditions. Users are invited to read the papers relative to the crops of interest and judge by themselves about the adequateness of the reported research.

Selection and tabulation of updated standard Kc and Kcb values

Standard values were established considering the ranges of Kc and Kcb values collected in the selected literature and the tabulated values since 1998. This work developed in the following steps:

First step: grouping the various studies relative to every crop considering:

  1. (i)

    The density of plants and spacing (DPS);

  2. (ii)

    The training and/or trellis systems used (TTS);

  3. (iii)

    The fraction of ground cover (fc), or fraction of the intercepted PAR (fIPAR); and

  4. (iv)

    The crop height (h).

Second step: building a provisional table for every crop. For all the groups of studies/papers, the ranges of Kc/Kcb ini, Kc/Kcb mid and Kc/Kcb end were defined and included as columns of Kc∕Kcb observed values in a provisional table relative to every crop. The ranges of previously tabulated Kc∕Kcb values in FAO56 (Allen et al. 1998), Allen and Pereira (2009), Jensen and Allen (2016), and Rallo et al. (2021) were also included as columns in that provisional table.

Third step: draft definition of the standard values for Kc/Kcb ini, Kc/Kcb mid and Kc/Kcb end for all crops through assessing the various ranges inscribed in each line of the provisional tables relative to sets of DPS, TT, fc or fIPAR, and h.

Fourth step: Definition of the standard values for Kcb ini, Kcb mid, and Kcb end for all crops through the computation of the A&P approach (Allen and Pereira 2009; Pereira et al. 2020b) for every set of fc and h using the parameters ML, which is a multiplier on fc describing the effect of canopy density on shading and on maximum relative ET per fraction of ground shaded [1.0–2.0], and Fr, which is is an adjustment factor relative to crop stomatal control [0.0–1.0]. ML and Fr are available in Pereira et al. (2021c) for most crops, or may be obtained by adjusting the parameter Fr for not previously validated values comparatively with similar crops.

Fifth step: Once defined the Kcb values, definition of the standard Kc by summing estimated values of Ke for each stage and the defined standard Kcb ini, Kcb mid and Kcb end. The estimated values of Ke were obtained from observing the differences (Kc-Kcb) in the selected papers and in the previously published Tables quoted above with consideration of changes in Kc due to rain and assuming a reduced soil evaporation due to using drip or micro-sprinkling under the canopies and/or for larger plant density. Young plantations are assigned with larger Ke values. Ke was assumed smaller for the mid-season, particularly for deciduous crops, and was also assumed smaller for the evergreen crops.

Sixth step: consolidating the draft standard Kc and Kcb through comparing all values (1) for various plant densities and ground cover fractions of the same crop; (2) for various crops of the same group, for instance within citrus; (3) for various training and trellis systems, e.g., among the multiple cases of grapes; and (4) between Kc and Kcb.

The Tables presenting the updated standard Kcb ini, Kcb mid, and Kcb end and standard Kc ini, Kc mid, and Kc end show their values in the last two columns, while the first ones are those indicating plant density and training or trellis systems, fc and h as well as the values assumed for ML and Fr relative to the initial, mid- and end-season stages, which may be useful for further uses of the A&P approach. The ranges of observed and previously tabulated Kc∕Kcb ini, Kc∕Kcb mid and Kc∕Kcb end are also included for information to users.

Table and wine grapes

Table and wine grapes are widely grown worldwide, even in less suitable environments, as they are the most popular woody Mediterranean crop. Domesticated after thousands of years, these plants have been cultivated for a long time using specific pruning and training systems that adapt to the climate of the site environment, the soil, the availability of water, the rootstock and the use of fruit (table grapes or wine). Furthermore, these variables also have impact on the timing of full bearing or maturity. Related knowledge is enormous, namely in relation to pruning and training as well as on water and irrigation requirements, including responses to the timing and severity of water deficits. Therefore, it has been possible to collect a variety of articles referring to evapotranspiration and crop water requirements aimed at irrigation scheduling. The collected information from the selected articles refer to the characteristics of the vineyards (Table 1) and to the observed crop coefficients (Table 2).

Table 1 Characteristics of selected vineyards
Full size table
Table 2 Field derived crop coefficients of table grapes and wine grapes vineyards
Full size table

The former studies for Kc tabulation in FAO56 (Allen et al. 1998) and by Rallo et al. (2021), and reporting for Tables relative to the use of the A&P approach (Allen and Pereira 2009; Pereira et al. 2020a, 2021c) have evidenced that abiotic factors—the fraction of ground cover by the crop vegetation (fc), which defines the amount of shadowed soil and the fraction (1−fc) from where soil water evaporates, the average crop height (h), the plants spacing or density in terms of plants per unit surface, the crop stress due to water or associate salinity, plant age, and training or trellis system—play a main role in determining crop evapotranspiration and transpiration. This is true, not only for grapes but also to other crops. Biotic impacts from the cultivar and rootstock are also to be considered as both influence the vigor of the crop. In addition, the destination of the grapes determines the dates of harvest and the Kc by the end of the crop season. The determination of the end season Kc results difficult when not appropriately defined, particularly because post-harvest irrigation may occur to provide for producing and store carbohydrates, less dry woody tissues, less incidence of winter injury, and to promote for an even bud break and shoot growth.

Collected values for characterizing the crop (Table 1) include these items, as well as the irrigation method and the irrigation strategy relative to water stress. A few of these factors are referred again in the Table 2, where Kc and Kcb are presented, which consist of the three values required to describe the traditional Kc and Kcb FAO curves.

Table 1 shows that selected studied vineyards cover a large, worldwide distribution of locations and cultivars for both table and wine grapes. This large coverage contributes to the desired perception of the reviewed case studies, hence providing for higher quality of the review. All reported actual Kc and Kcb were derived from field surveys of ETc act, which used a large panoply of measurement methods with verifiable accuracy. ETo has been always computed with the FAO-PM-ETo equation or similar, while ETc act has been observed with WL and DL lysimeters, or computed with SWB from soil profiles or DL observations, in some cases using the SIMDualKc model for data handling, BREB and EC measurements of energy balance, or the SR approach and observations to use the A&P approach. In addition, there are various cases relative to obtaining Tc act from SF measurements.

All vineyards but one were micro-irrigated adopting a full irrigation strategy. Drip was largely the main irrigated method and just two cases used micro-sprinkling. Various types of drippers were used. These conditions prefigure good control of water application depths and low soil evaporation. The sole non-irrigated vineyard is located in the sub-alpine slopes of Alessandria, in Italy, where precipitation is enough to satisfy crop water requirements. It could be assumed that the selected papers report on vineyards that have not been under water stress except for short periods due to deficient irrigation scheduling, thus corresponding to the conditions defined for standard crop coefficients. As per Table 2, most cases refer to bare soil (BS) and only a few to active ground cover (AGC), generally during and immediately after the rainy season, which require a specific solution to identify ET from the crop and from the AGC, and soil evaporation (Rosa et al. 2012a, b).

There is a great variability of spacing and planting densities, that relate with the trellis system used, and a great variety of training and trellis systems, which, generally, are differentiated into the two groups of table grapes and wine grapes. Reported information on pruning was very scarce and, therefore, is not referred herein; however, some studies reported that pruning was performed annually during dormancy, in a few cases also a slight summer pruning, depending upon the vigour of the plants. The diverse training and trellis systems determine crop height and the fraction of ground cover, with fIPAR assumed as an estimate of fc. Collected data confirms that fc and h are generally larger for table grapes than for wine grapes. However, there is a very large variability of fc data for both table and wine vineyards. That variability also relates with age, with young plants (< 5 years) having smaller fc. But the variability of fc also refers to crop conditions and age that favor ground shadow, e.g., the cases studied by López-Urrea et al. (2012) and Picón-Toro et al. (2012) that show a correspondence between fc and Kc mid or Kcb mid.

The trellis systems for table grapes vineyards are dominantly overhead trellis, “Y” or “T” trellis, cross-arm trellis and high vertical shoot positioning (VSP), thus resulting in h ≥ 2.0 m. For wine grapes, overhead systems (e.g., pergola) are rare and a variety of trellis systems are used such as VSP, single and double Guyot, single and bilateral cordon, Y-trellis bilateral cordon, Guyot, Lyre trellis, GDC trellis, and QCT (Quadrilateral cordon trained). Detailed descriptions of trellis systems and their relations to cultivars and vineyards mechanization were given by Fidelibus (2014), and an analysis of relationships between trellis systems, shot positioning, and light interception is available in Louarn et al. (2008). Wider analysis referring to trellis systems, canopy architecture, water use, and Kc values is provided by Williams and Fidelibus (2016) and Williams et al. (2022).

Actual Kc and Kcb values obtained from field ET observations and the corresponding ETo values are presented in Table 2 for all reported cultivars and rootstocks together with factors that mainly influenced them: fc, h, and trellis systems. Tabulated observed Kc mid and Kcb mid show to vary greatly among cultivars and for the same cultivar, as well as with the trellis system. Finding the most appropriate standard Kc and Kcb values would be nearly impossible without following the conclusions of Williams et al. (2022) that the prime factor influencing Kc values is training and trellis system disregarding if vineyards are of table or wine grapes. It resulted then a good organization of study results and the Table 3 was then built, first a draft working approach as referred in Section“Selection and tabulation of updated standard Kc and Kcb values”, then simplified as shown herein. Plant density will vary within a given training system and, therefore, their ranges of values in Table 3 shall be considered as indicative to users.

Table 3 Initial, mid-, and end-season standard single and basal crop coefficients for vineyards as related with the training trellis system, fraction of ground cover and height for table and wine grapes with indication of ranges of observed Kc and Kcb, and of former tabulations of their standard values
Full size table

Table 3 shows initial, mid-, and end-season Kc and Kcb of table and wine grapes grouped according to the degree of ground cover (DGC), training and trellis system (TTS), and plant density and spacing (PDS). DGC varies from very low when plants are young (< 5 years), to high in case of overhead trellis in table grapes, or to very high in case of well covering Y-trellis, the Geneva Double Curtain double wire system, and the overhead trellis system. The diverse degree of ground cover corresponds to diverse TTS, which are influenced by the pruning intensity, and to various plant density and spacing. The described groups are also characterized by ranges of the fraction of ground cover and height, fc and h, which may help to decide which group is more appropriate for the case under study. Moreover, fc and h may be utilized to compute Kcb for the three stages with the A&P approach (Allen and Pereira 2009; Pereira et al. 2020a) with help of the proposed parameters ML and Fr, also tabulated.

The proposed standard Kcb and Kc are given in the last two columns of the Table 3. The ranges of Kcb and Kc obtained from field measurements and proposed in the selected papers and the ranges reported in previous Tables, namely FAO56, are also tabled as they were used for selecting the values of the proposed standard coefficients. Readers are advised to interpolate the proposed Kcb and Kc using the data they have available.

It is evidenced by Table 3 that standard Kcb and Kc for vineyards mainly increase with ground cover and plant density, thus depending upon training and trellis systems as they favor or not ground shading, thus the light intercepted by the canopy. Soil evaporation, contrarily, is governed by the TTS that provide for larger or limited solar radiation at the soil surface, thus for larger or reduced energy for soil water evaporation.

Olive orchards

Olive trees are, after centuries, main references of the Mediterranean landscapes, either isolated or in small groups, or in orchards. Due to their physiological characteristics, olives are resistant to dryness and droughts and other abiotic stresses (Fernández 2014) but then decreasing growth and yield. Climate change is affecting olives water requirements and, then, the landscape (Tanasijevic et al. 2014). Traditional orchards are rainfed, have wide tree spacing, and are vase-trained. They continue to be used but are declining and being replaced by irrigated orchards with increasing plant density, such as the modern super high density hedgerow system. Plant density increased from 225–250 trees ha−1 to almost 2000 trees ha−1. Unlike vineyards, fc and h show a relatively little variability, however with exceptions.

The selected orchards (Table 4) are mainly located in the Mediterranean region, with only one from a Chilean location with Mediterranean climate, the Talca Valley. Cultivars are often changing from traditional ones (e.g., ‘Picual’, ‘Cobrançosa’) to cultivars adapted to high density systems like ‘Arbequina’. Training systems and plant density affect tree maturity, with intensive orchards trained in vase, reaching full bearing by 7–8 years, while high-density systems, trained as hedges, reach full bearing 4–5 years after planting. In these latter systems, the mechanical harvesters limit plant height and therefore pruning is mandatory at least once a year.

Table 4 Characteristics of the selected olive orchards (Olea europea L.)
Full size table

All reported ETo computations refer to the FAO-PM-ETo. Field ET studies were performed mostly using EC systems and SWB with various sensors and the model SIMDualKc; SF systems were used for measuring transpiration. Drip irrigation was used in most cases but always adopting controlled or regulated deficit irrigation (Table 4). The eustress concept used for vineyards does not apply to olives due to their resistance and resilience to droughts and water stress, which calls to adopt limited water use and costs by currently adopting deficit irrigation, mainly during the pit hardening stage.

The crop coefficients show some variability, both Kc and Kcb (Table 5). The Kc curves present a mid-season value lower than the Kc ini and Kc end which is the consequence of the Mediterranean rainfall regime, with rain by the initial stage, by the early spring, and by the final stage, at mid-autumn, with a dry summer mid-season. The Kcb curves are different because transpiration is much higher in mid-summer under irrigation than during the initial and end stages. The Med climate does not change much inter-annually, but global change is making the dry summer season longer. Without irrigation both Kc and Kcb curves tend to flat down due to impacts of water stress. Soil evaporation is important during the non-growing period, mostly the winter, when precipitation occurs; it is negligible during the mid-season if rain is very low and irrigation is under canopy as for drip and micro-sprinkling; intermediate conditions occur by the initial and final periods depending upon the distribution of rainfall. AGC and mulch are rarely practiced but natural AGC occurs during spring, with AGC converting into residual mulch during the summer.

Table 5 Field derived crop coefficients of olive orchards (Olea europea L.)
Full size table

The proposed standard Kcb values in Table 6 were defined in agreement with the observed ranges, often slightly larger than these. But the Kc values are smaller than the observed ranges and the ranges previously tabulated because when irrigation is under the canopies and practiced with good quality equipment and efficiently there are no reasons for high soil evaporation or operational losses. This is particularly true in the hedgerow olive orchards since differences between ranges observed for Kcb and Kc are high, likely indicating non negligible soil evaporation losses. As for vineyards, the plant densities referred in Table 6 are guidelines for users.

Table 6 Initial, mid, and end-season standard single and basal crop coefficients for olive orchards as related with the training trellis system, fraction of ground cover and height for table and wine grapes with indication of ranges of observed Kc and Kcb, and of former tabulations of their standard values
Full size table

Citrus orchards

Various tree species are included among the citrus trees: clementine, grapefruit, lemon, lime, mandarin, and orange. Studies relative to orange are by far the most common, followed by clementine and mandarin. Studies were carried mostly in the Mediterranean region but those for lime (Citrus latifolia Tan.) were developed in Brazil, where this crop is very popular; those for orange, following its wide dissemination, in addition to the Med basin, come from North and South America, South Africa and Iran (Table 7). This wide origin of the selected studies proposes various perspectives that favor the analysis aimed at finding the appropriate standard Kc and Kcb for all the crops. Moreover, cultivars referred for each crop are also diverse.

Table 7 Characteristics of the selected studies on citrus orchards
Full size table

As for the grapes and olives, the appropriateness and accuracy of computation of ETo and crop evapotranspiration were analyzed (Table 7). In most cases, the FAO-PM ETo was adopted with only 2 cases of using the Penman equation and the class A pan were observed. A diverse panoply of field ET measurement was reported. The most common approach was SWB based on diverse soil water sensors, followed by the sap-flow measurement of transpiration and the EC measurement of ET, often combined. Weighing lysimeters and the surface renewal method were also used. Drip and micro-sprinkler irrigating under canopy were generally adopted. Full irrigation, sometimes in excess, was the main strategy. Information reported in literature was however insufficient to understand if, likely, eustress was considered. In general, it could be considered that conditions existed to favor water saving and high yields.

A variety of plant densities and spacing are reported but it was not possible to relate them to training, with many papers not reporting about training. The most common is vase but some hedgerow, yet with relatively low plant spacing, were also referred. Generally, plant heights varied from about 2.5 to 4.0 m but much larger and uncommon heights near 6 m were reported for orange in Florida. Tree height was lower for mandarin (< 2.8 m) and clementine (< 4.1 m). The fraction fc followed a similar trend, smaller for mandarin, lime, lemon and clementine, largest for orange (up to 0.90). Differences in architecture and sizes, as well as physiological but not referred herein, justify that Kc and Kcb were not given in a single group of citrus.

Generally, results in Table 8 show well the dependence of Kc mid and Kcb mid on crop age, height, and fc as it is the case of studies by Castel (2000) for clementines, Alves et al. (2007) for lime trees, Maestre-Valero et al. (2017) for mandarin, and Consoli et al. (2006) for orange. Since citrus are evergreen trees, they also show a Kc curve where higher values are for Kc ini and Kc end for climates like Mediterranean, with very small precipitation in summer and the rainfall season initiating by the fall and ending by the spring. In other climates, this may not happen. Because citrus are evergreen and for some species or cultivars show differences in crop stages, some growers and advisers adopt a constant Kc, which lead to flat down the period between spring and winter, i.e., when irrigation is required. Several citrus studies report Kc/Kcb values on a monthly time scale, so the growth stages of the plants were defined according to the tree’s annual cycle (3 vegetative growth peaks corresponding to spring, summer and autumn). Therefore, the initial stage corresponds to flower initiation (December–January in the Northern Hemisphere, June–July in the Southern Hemisphere), the mid-season stage is a very long period corresponding to fruit growth extending from March to November (Northern Hemisphere) or from September to May (Southern Hemisphere). The end-season occurs after maturation and harvesting, i.e., in November (Northern Hemisphere) or May (Southern Hemisphere). However, these stages depend on the species and cultivar. It is, therefore, advisable to define well the initial, crop development, flowering and fruiting mid-season, and maturation and harvesting. Then the Kc and Kcb curves are expected to be as referred above, however distinct among species and, less, cultivars (Table 8).

Table 8 Field-derived crop coefficients of citrus orchards
Full size table

The definition of the standard Kc and Kcb (Table 9) followed the same methodology used and shown for grapes and olives. Initially, all citrus trees were considered together but, due to differences among the various species, three groups were considered. Thus, clementine, lime and mandarin trees consist of the first group of species, which is characterized by the smaller tree height, fraction of ground cover and Kc/Kcb for each degree of ground cover, training, and plant density. Orange, grapefruit, and tangelo, in contrast, have trees with higher h, fc and Kc/Kcb values. Lemon trees are in an intermediate position. Nevertheless, the generally required interpolation may be difficult.

Table 9 Initial, mid-, and end-season standard single and basal crop coefficients for citrus orchards as related with the training system, fraction of ground cover and height, with indication of ranges of observed Kc and Kcb, and of former tabulations of their standard values
Full size table

Warm temperate plantations: avocado, loquat, persimmon, and tea crops

These crops are not grouped but listed in the same Table. They have great differences: on the one hand, they are evergreen but persimmon that is deciduous; on the other hand, all are trees explored for fruits but tea, which is a shrub explored for the leaves. Thus, tabulated subjects are discussed in isolation or comparatively.

Selected studies on avocado orchards are from Florida, South Africa and Chile, which are among the main producers (Table 10). Only recently, they start to be grown in southern Europe, which may be a consequence of global warming as suggested in a review by Cárceles Rodríguez et al. (2023). Differently, loquat and persimmon have long been cultivated in southern Europe and the selected studies are from the north and east of the Mediterranean region. The selected tea studies are from two main production areas, southern China, and mountainous India, but tea has a quite large distribution, which is also related to the qualities of tea produced.

Table 10 Characteristics of selected avocado, loquat, persimmon, and tea plantations
Full size table

The FAO-PM-ETo was adopted for most studies on the various crops (Table 10). Field ET measurements with a SWB approach, followed by EC systems, were the main methods used for avocado ET estimation. For persimmon, EC systems were the main methods to measure ET. Differently, for loquat a test of Kc fitting was employed.

The planting density reported for avocado (Table 10) ranged 148–370 pl ha−1 and the training systems reported were hedge pruned or, more often, hedgerow. However, these systems are very different of those used for olive trees since crop heights are quite high, of up to 7.9 m. These hedge systems aim to improve harvesting efficiency, which occurs throughout the year, using Harvest Assist platforms. A large range of heights results in a wide range of fc values, from 0.40 to 0.80. Persimmon and loquat have training in vase while tea is trained at a low hedgerow, with h < 0.90 m, to favor hand harvesting of leaves.

Crop coefficients of avocado are reported with two types of Kc curves (Table 11): where the summer mid-season is dry in opposition to the initial stage and the final stage, the Kc curve has Kc mid smaller than Kc ini and Kc end because mid-season soil evaporation is about negligible; if there is rain in the mid-summer, it is likely that soil evaporation is high by then resulting a Kc mid higher than Kc ini and Kc end. The difference among these Kc may be small, then resulting a uniform season Kc. Kcb is reported to follow a typical segmented crop coefficient curve with Kc mid higher than Kc ini and Kc end assuming that transpiration is larger during the mid-season stage; however, differences among these Kcb values may be small as it often happens to citrus trees. However, for the New Zealand case (Kaneko et al. 2022), with observations in three different locations, Kcb end > Kcb mid is reported but without explanations.

Table 11 Field-derived crop coefficients of avocado, loquat, persimmon, and tea plantations
Full size table

There is limited information about loquat (Table 11) but it is likely, as reported by Hueso and Cuevas (2010), that Kc mid and Kcb mid be larger than initial and end-season values. On the one hand, flowering occurs by the end of winter and fruit maturation is also anticipated to the spring, thus the crop mid-season is likely when rainfall occurs, resulting that transpiration adds to non-negligeable soil evaporation due to rains occurring by then, thus with Kc resulting from the sum of Kcb mid with a non-negligible Ke value. On the other hand, leaves are tough and leathery in texture, and densely velvety-hairy below that favor stomatal control during the late season.

Despite data are limited, reports show that Persimmon has Kc and Kcb curves with the mid-season values larger than the initial and end season values, as it is common for deciduous trees.

Reported tea results for Kc and Kcb (Table 11) show flat Kc and Kcb curves since the climate where plantations develop is generally humid, with only short dry spells, which does not favor ET values very different of those of the grass reference, therefore close or equal to 1.0.

The proposed initial, mid-season, and end-season standard single and basal crop coefficients for avocado, loquat, persimmon, and tea plantations (Table 12) are generally in agreement with the ranges of values observed and compatible with those previously tabulated for avocado and tea. However, hedgerow was not yet considered previously for avocado while presently it is likely the most popular where harvesting mechanization is in use; nevertheless, training in vase is continued. Kc for loquat and for persimmon were never tabulated. Proposed values for these crops agree with previous discussions. As previously pointed out the tabulated values of the ranges of plant densities are indicative.

Table 12 Initial, mid-, and end-season standard single and basal crop coefficients of avocado, loquat, persimmon, and tea plantations as related with the training system, fraction of ground cover and with indication of ranges of observed Kc and Kcb, and of former tabulations of standard values
Full size table

Conclusions and recommendations

The review of crop coefficients for table and vine grape vineyards, olive, citrus, avocado, loquat, persimmon, and tea plantations permitted a good collection of well-performed field studies and data handling that elucidated about water use practices and requirements for those crops. The selected papers allow to conclude that good knowledge exist about the referred crops and their exploitation, the evapotranspiration and water use process, while water management practices require to be improved in such a manner that water use be controlled, limited, while yields are increased. However, further studies on crops having limited information available are welcome, e.g., lemon and loquat.

The control and optimization of water use, including water saving, require appropriate choice and use of irrigation equipment and adequate irrigation scheduling targeting the standard Kc when irrigation equipment allows a good control of quantities applied and water available is enough to satisfy that target application. Numerous papers refer to regulated or controlled deficit irrigation; however, that deficit must be referred to the potential ETc, product of the standard Kc by the grass reference ETo. The application of those deficit irrigation practices also imply that farmers, managers and farmer advisers improve their knowledge on these subjects, on using models that may help decision-making, as well as on the use of weather data and information. Estimating Kcb and Kc from the fraction of ground cover or shading and plant height (A&P approach) provides for quite realistic estimates of crop coefficients for trees and vines as demonstrated by Pereira et al. (2020a, b) and as used with the California remote sensing SIMS framework (Melton et al. 2012; Pereira et al. 2021a, b, c), with parameterization described by the latter. Similar approaches on the use of standard Kc and Kcb, or the A&P approach, apply to studies on irrigation planning as well as on consumptive use assessment at project or watershed level.

When searching for water saving and scheduling irrigations for any kind of controlled deficit irrigation, users may either use the standard Kc or Kcb decreased by a saving fraction or may schedule irrigations following the actual ET conditions of the orchard. In the latter case, users may estimate the Kc act using the A&P approach as referred in the Introduction. Then Kcb act may be computed from the observed actual fraction of ground cover and crop height and Ke may be estimated from the observed actual wetted fraction of exposed soil, 1−fc. The resulting actual values for Kc or Kcb shall then be compared with the standard Kc or Kcb for computation quality control. It is important to make the best use of related information and effectively achieve high water and financial productivity aiming the sustainability of production and the progressive adaptation to climate change challenges.

Users are advised to read and analyze the quoted papers in addition to the information provided and tabulated in the current review paper. Above all, it is required to develop awareness on water scarcity and water saving, the latter mainly based on the knowledge of standard crop coefficients and related transfer for different locations and diverse climate conditions.