Analysis of on-farm irrigation performance inMediterranean greenhouses

M.D. Fernandez a,*, A.M. Gonzalez a, J. Carreno a, C. Perez a, S. Bonachela b

aEstacion Experimental de la Fundacion Cajamar, Autovıa del Mediterraneo, km. 416.7, 04710 El Ejido, Almerıa, SpainbDepartamento de Produccion Vegetal, Universidad de Almerıa, La Canada de San Urbano s/n, 04120 Almerıa, Spain

a g r i c u l t u r a l w a t e r m a n a g e m e n t 8 9 ( 2 0 0 7 ) 2 5 1 – 2 6 0

avai lab le at www.sc iencedi rec t .com

journal homepage: www.e lsev ier .com/ locate /agwat

a r t i c l e i n f o

Article history:

Accepted 7 February 2007

Keywords:

Water use

Drip irrigation

Horticultural crops

Performance indicators

Greenhouse

Mediterrranean

a b s t r a c t

A comprehensive irrigation assessment was conducted using on-farm water use informa-

tion and simulated crop water requirements in a Mediterranean greenhouse area, mainly

dedicated to horticultural crops, located on the Almerıa coast.

The mean irrigation water supply (IWS) for the main greenhouse crop cycles was 228 mm

and ranged from 158 mm (autumn green bean) to 362 mm (autumn–spring sweet pepper).

Besides, the mean AIWS value for the main crop rotations was 444 mm and ranged between

363 mm for autumn–spring sweet pepper and 502 mm for autumn–winter pepper and spring

melon.

Mean relative irrigation supply (RIS) values were close to 1 for most greenhouse vegetable

crops, indicating that, on average, the irrigation supply matched the maximum water

requirements of these crops. By contrast, the mean RIS value of autumn–winter cucumber

was 1.6, indicating that, on average, the irrigation supply clearly exceeded the calculated

optima. However, for most crops, the high CV values observed for RIS and the analysis of the

RIS dynamics throughout the cycles indicate that there are greenhouse crops and crop cycle

periods for which the IWS clearly does not match the crop water requirements. Greenhouse

irrigation water use in the Almerıa coastal region can, therefore, be improved.

Mean irrigation water use efficiency (IWUE) values for greenhouse horticultural crops

ranged from 15.3 kg m�3 (autumn–winter green bean) to 35.6 kg m�3 (spring watermelon).

They were, in general, higher than those found when these crops were grown outdoors in

similar climatic regions. Water productivity (WP) varied from 7.8 to 15.9 s m�3 and were

highest for green bean crops. WP values of greenhouse crops were generally much higher

than those found in other irrigation districts around the world, including Mediterranean

areas, due to the low IWS and, especially, to the high value of the vegetable crops grown

off-season.

# 2007 Elsevier B.V. All rights reserved.

1. Introduction

Greenhouse vegetable production is expanding in many world

regions (Enoch and Enoch, 1999) and in particular throughout

Mediterranean coastal areas (Pardossi et al., 2004). The

Mediterranean greenhouse vegetable system is mostly based

* Corresponding author. Tel.: +34 950 580 569; fax: +34 950 580 450.E-mail address: mdfernandez@cajamar.es (M.D. Fernandez).

0378-3774/$ – see front matter # 2007 Elsevier B.V. All rights reservedoi:10.1016/j.agwat.2007.02.001

on simple low technology plastic greenhouses located in mild

winter areas, which enables the production of high-value

vegetables from autumn to spring (Castilla and Hernandez,

2005).

Growing populations and an expected higher standard of

living will increase water demand dramatically in the near

d.

Fig. 1 – Distribution of the monitored greenhouses in the

Sol y Arena (SAID) and the Sol-Poniente irrigation districts

(SPID), and from small irrigation districts (SID) located on

the Almerıa coast, southeast Spain.

a g r i c u l t u r a l w a t e r m a n a g e m e n t 8 9 ( 2 0 0 7 ) 2 5 1 – 2 6 0252

future. Irrigation worldwide (including Spain) accounts for

about two thirds of all water usage, and so there are

increasing societal demands for a more productive use of

this resource (Howell, 2001) and for an effective account-

ability of irrigation water use (Seckler, 1996). Therefore,

performance evaluation of irrigation areas is needed to

propose improvements in irrigation management, to assess

water productivity and to preserve the environment (Molden

and Sakthivadivel, 1999; Lorite et al., 2004a). In semiarid

Mediterranean coastal areas with increasing water demands,

such as the Almerıa coast, where intensive agriculture,

tourism and other sectors are competing for scarce water

resources, evaluation of irrigation performance including

water productivity is going to be a prerequisite for any future

water policy.

The analysis of irrigation performance is usually conducted

with a set of indicators (Molden and Gates, 1990; Malano and

Burton, 2000), which could be locally adapted to account for

the idiosyncrasies of each irrigation area, so that meaningful

assessments may be carried out (Lorite et al., 2004b). Scheme-

level assessments, producing performance indicators based

on average values for the whole irrigation area (Molden et al.,

1998), are the only feasible approach when information at sub-

scheme-levels (usually farmers) is not available. However, this

procedure may not reflect accurately the current irrigation

practices in the area, since it does not measure the degree of

variation in irrigation management among farmers. The

availability of water-use information at the individual plot

or farmer level allows in-depth assessment of irrigation

performance by characterising average performance indica-

tors on each of the main crops in the area and the variation

among farmers for any of these indicators. High variability in

performance among farmers would indicate a substantial

potential for improvement, even if average performance

values are reasonable. In a further step, a more detailed study

could be focussed on farmers with performance values outside

the reasonable range.

Irrigation performance studies have already been con-

ducted in several semiarid areas dedicated to open field crops

(Faci et al., 2000; Dechmi et al., 2003; Lorite et al., 2004a,b;

Lecina et al., 2005), but not in greenhouse areas, which

usually present rather different characteristics in soils

(artificial soils), crops (intensive crops), irrigation practices

and socio-economic conditions (high water cost, high-value

crops, etc.). The greenhouse area located on the Almerıa

coast, one of the largest in the world, is drip irrigated and

mainly dedicated to horticultural crops. Despite the rela-

tively high irrigation water prices, current irrigation practices

are generally based on local growers’ experience (soil or plant

water sensors are not normally used) and high variations in

irrigation water supplies to each of the main vegetable crops

have recently been detected (Caja Rural de Almerıa, 1997;

Gonzalez, 2003). Additionally, some water contamination

problems of the underlying aquifers have been already

detected (Pulido-Bosch et al., 2000). The aim of this work was

to conduct a comprehensive assessment of the irrigation

performance in this area using on-farm water use informa-

tion and simulated crop water requirements in order to

improve the irrigation water management and minimise

percolation losses.

2. Materials and methods

2.1. Area description

The study area was located within the Campo de Dalıas, on the

west coast of the Almerıa province, in southeast Spain (Fig. 1).

This is the largest and oldest greenhouse area on the

Mediterranean Spanish coast with approximately 20,500 ha

of plastic greenhouses (Sanjuan, 2004), mainly dedicated to

vegetable production. The climate is Mediterranean with mild

winters and low annual precipitation: average annual tem-

perature and rainfall are 18 8C and 220 mm, respectively.

Artificial layered soils, locally called ‘‘enarenados’’ (Wittwer

and Castilla, 1995), are mostly used by greenhouse farmers,

although inert substrates (perlite and rockwool) in plastic bags

are being used increasingly. Greenhouses are mostly low-cost

structures covered with plastic film, called ‘‘Parral’’ (Perez

Parra et al., 2004), located on practically flat plots. On average,

they are 10 years old and have a surface of 7300 m2. The main

crops in the Campo de Dalıas area were pepper, cucumber,

green bean, melon and watermelon.

There are more than 100 water irrigation districts in the

Campo de Dalıas (Caja Rural de Almerıa, 1997). The evaluation

study was conducted in the two main irrigation districts,

called Sol y Arena (SAID) and Sol-Poniente (SPID), respectively,

and in a representative group of greenhouses supplied from

small irrigation districts (SID). Water of relatively good

irrigation quality [electrical conductivity (EC) values usually

about 1 dS m�1, but always lower than 2 dS m�1] and provided

by deep wells and a nearby reservoir was mostly used. The

water distribution system in the SAID and SID consisted of a

gravity-fed branched network, mostly concrete irrigation

ditches, from which water is diverted weekly to farmers’

irrigation ponds (Caja Rural de Almerıa, 1997). In the SPID the

water was distributed by a pressurized irrigation system,

which allows flexibility of frequency and duration of water

delivery to each greenhouse. Most farmers have small

reservoirs or ponds close to the greenhouses for ensuring

a g r i c u l t u r a l w a t e r m a n a g e m e n t 8 9 ( 2 0 0 7 ) 2 5 1 – 2 6 0 253

water availability, as these high-value crops are usually

irrigated frequently. All greenhouses are drip irrigated, and

the irrigation head, usually near the ponds, consisted of a

pump, water filters, fertilizer injector and control mechan-

isms. Water is distributed within greenhouses by polyethylene

drip lines with emitters of nominal discharge rates ranging

from 2 to 4 L h�1 placed every 0.5 m.

2.2. Data collection

The study was carried out over six cropping seasons (1993/

1994–1998/1999) in 41 representative greenhouses chosen by

local experts from the SAID (58 monitorised greenhouse data),

the SPID (80) and the smaller irrigation districts (62). The

spatial study unit was the greenhouse. Total irrigation water

supply per crop (IWS, mm) and cropping season (AIWS, mm)

were measured by water-meter readings installed in the main

irrigation pipe entering each greenhouse. Measurements were

taken approximately every month and at the start and end of

each crop cycle. Information on plot size (greenhouse), crop

and cycle, variety, sowing or planting date and plant

population was obtained from personal interviews with each

greenhouse farmer. They also provided data of total fresh fruit

production (CP, kg m�2) and gross income (s m�2) per crop

cycle and greenhouse. Characteristics of each drip irrigation

system (emission uniformity, EU) were measured at the

beginning of the study (Merriam and Keller, 1978), while the

irrigation water electrical conductivity (ECw) was determined

once per cropping season. Average ECw and EU values for the

whole period and area studied were 1.05 dS m�1 and 91%,

respectively. These variables were measured at the beginning

of the cropping season.

2.3. Simulation model

Crop water requirements were calculated as the maximum

crop evapotranspiration (ETc) plus the additional water

required for leaching of salts or for compensating for non-

uniformity of water application (Allen et al., 1998). Leaching

requirements were calculated using EC irrigation water data

(Ayers and Westcot, 1985).

The effective rainfall was considered practically null, as the

amount of rainwater entering the greenhouse is usually small

and heterogeneously distributed. Thus, the irrigation water

requirements were equal to the crop water requirements. ETc

was calculated using the Kc–ETo method recommended by the

FAO (Doorenbos and Pruitt, 1977; Allen et al., 1998) and

adapted to greenhouse crops by Orgaz et al. (2005):

ETc ¼ ETo � Kc (1)

The reference crop evapotranspiration (ETo) was calculated

with a locally calibrated radiation method (Bonachela et al.,

2006) that requires daily solar radiation data and greenhouse

transmissivity estimates. Kc values for major greenhouse

horticultural crops were determined according to Fernandez

(2000) and Orgaz et al. (2005). Daily Kc values from sowing/

planting to effective full cover were determined as a function

of thermal time, calculated using greenhouse air temperature.

Real-time meteorological data required for determining ETc

values were taken from a weather station located within the

‘‘Cajamar Foundation’’ research station (28430W; 368480E; and

155 m a.s.l.) in the Campo de Dalıas. Greenhouse air

temperatures were measured within a typical Mediterranean

greenhouse located in the same research station.

Based on local practices and their own experience, green-

house farmers usually irrigate abundantly before crop sowing

or planting, and frequently later on (daily in periods of high

water demand). The soil water content is not measured or

considered in the greenhouse water balance. Farmers usually

try to maintain the soil water content around field capacity.

2.4. Irrigation performance indicators

Two irrigation performance indicators (Malano and Burton,

2000) were used: relative irrigation supply (RIS) and irrigation

water productivity (WP) per crop cycle. These indicators were

also determined annually (ARIS and AWP, respectively), as

many farmers usually cultivate two horticultural crops per

season. The irrigation water use efficiency (IWUE) was also

assessed (Sinclair et al., 1984). For greenhouse crops, the RIS

indicator is equal to the relative water supply indicator

(Malano and Burton, 2000) since all the water requirements

were applied by irrigation, because the effective rainwater

entering the greenhouse was considered negligible. We

defined the indicators as:

RIS ¼ total crop irrigation water supply ðIWSÞtotal irrigation water requirements ðIWRÞ (2)

ARIS ¼ annual irrigation water supply ðAIWSÞannual irrigation water requirements ðAWRÞ (3)

WP ðs m�3Þ ¼ total value of crop productiontotal crop irrigation water supply ðIWSÞ (4)

AWP ðs m�3Þ ¼ annual value of crop productionannual irrigation water supply ðAIWSÞ (5)

IWUE ðkg m�3Þ ¼ marketable crop productiontotal crop irrigation water supply ðIWSÞ (6)

3. Results and discussion

3.1. Irrigation district comparison

Irrigation performance indicators for the three irrigation

districts studied are presented in Table 1. These values are

the averages of six cropping seasons and include the major

crop cycles of each irrigation district: autumn–winter cycles of

sweet pepper, cucumber and green bean, spring cycles of

melon, watermelon and green bean, and autumn–spring

sweet pepper. No significant differences were found between

irrigation districts for IWS nor for any of the irrigation

performance indicators (RIS, IWUE and WP), although the

ECw was higher in the SAID than in the SPID and SDI.

Greenhouse farmers with irrigation water of relatively high

ECw from the SAID either reduced slightly the concentration of

Table 1 – Mean values of emission uniformity (EU), irrigation water electrical conductivity (ECw), irrigation water supply(IWS), relative irrigation supply (RIS), irrigation water use efficiency (IWUE) and irrigation water productivity (WP) per cropin the irrigation districts of Sol y Arena (SAID) and Sol-Poniente (SPID), and in the greenhouse group from small irrigationdistricts (SID). (Almerıa, southeast Spain)

Irrigation district Na EU (%) ECw (dS m�1) IWS (mm) RIS IWUE (kg m�3) WP (s m�3)

SAID 58 91 1.44 b 245 1.15 25.4 11.8

SPID 80 95 0.90 a 226 1.16 25.7 11.3

SID 62 86 0.70 a 216 1.06 20.7 12.8

Values with different letters in the same column are significantly different (P < 0.05).a Number of measured greenhouse crops per irrigation district.

a g r i c u l t u r a l w a t e r m a n a g e m e n t 8 9 ( 2 0 0 7 ) 2 5 1 – 2 6 0254

the applied nutrient solution or grew vegetable crops less

sensitive to salinity (e.g. green bean was scarcely cultivated in

this irrigation district).

Data of irrigation performance indicators were previously

analysed per crop cycle and, in general, no significant

differences were found between irrigation districts (data not

shown). Consequently, the evaluation of the Campo de Dalıas

irrigation area was carried out considering data from the

different irrigation districts as a whole. Overall, the mean

irrigation depth per crop cycle was 228 mm, which slightly

exceeded the maximum irrigation water requirement,

whereas the mean water productivity was 11.9 s m�3.

3.2. Irrigation water use

Total irrigation water applied to the main greenhouse crop

cycles (IWS), averaged over the six cropping seasons studied,

ranged from 158 mm (green bean) to 311 mm (pepper) for the

autumn–winter cycles and from 177 mm (melon) to 197 mm

(green bean) for spring cycles, and 363 mm for the autumn–

spring pepper (Table 2). IWS values increased with the crop

cycle length, being lower for green bean, melon, cucumber and

watermelon crops (3–4 months) and higher for autumn–winter

and autumn–spring pepper crops (5–9 months). For regions of

Table 2 – Values, averaged for six cropping seasons and coeffic(IWS) per greenhouse crop cycle in the irrigation area of Campofrom Mediterranean areas

Crops Campo de Dalıas

Na IWS (mm)

Autumn–winter pepper 33 311 (32)

Autumn–winter cucumber 38 270 (40)

Autumn–winter green bean 21 158 (33)

Spring melon 53 177 (31)

Spring watermelon 22 189 (38)

Spring green bean 22 197 (24)

Autumn–spring pepper 11 363 (30)

a Number of measured greenhouse crops per crop cycle.b Significant amounts of rainfall during the crop cycle.

similar climate, IWS values of horticultural crops were much

lower when they were grown in greenhouses than outdoors

(Table 2; Moller and Assouline, 2007), except when field crops

received a significant part of their water requirements via

rainfall (Erdem and Yuksel, 2003; Orta et al., 2003). This was

due to the comparatively lower evaporative demand inside the

greenhouse: lower solar radiation and wind in the greenhouse

(Montero et al., 1985; Moller and Assouline, 2007); a further

reduction in solar radiation by cover whitening in spring,

summer and early autumn (Castilla and Hernandez, 2005); and

crops growing mostly during low evaporative demand periods

(autumn, winter and spring; Orgaz et al., 2005). There was

substantial variability in the 6-year IWS values with coeffi-

cients of variation (CV) oscillating between 24% for spring

green bean and 40% for cucumber (Table 2). Part of this IWS

variation could be attributed to the wide range of crop cycles

(i.e. autumn–winter sweet pepper usually starts from June

until October and finishes from December to March).

In addition, average 6-year values of simulated irrigation

water requirements for the studied greenhouse vegetable

crops (data not shown) ranged from 140 mm (autumn–winter

green bean) to 362 mm (autumn–spring pepper). These

values were 11% higher than the corresponding simulated

values of crop evapotranspiration, due to the high water

ients of variation (in parenthesis) of irrigation water supplyde Dalıas and comparative IWS values of open-field crops

Open field

IWS (mm) Country References

496 Spain Moreno et al. (2000)

570 Turkey Sezen et al. (2006)

844 Turkey Simsek et al. (2005)

509 Turkey Ertek et al. (2006)

407 Spain Fabeiro et al. (2002)

152b Turkey Erdem and Yuksel, 2003

241b Turkey Orta et al. (2003)

737 Turkey Simsek et al. (2004)

340 Turkey Sezen et al. (2005)

366b Turkey Gencogla˘n et al. (2006)

Table 3 – Values, averaged for six cropping seasons, and their coefficients of variation (in parentheses) of annual irrigationwater supply (AIWS), annual relative water supply (ARIS) and annual water productivity (AWP) of the main crop rotationsin the irrigation area of Campo de Dalıas (Almerıa), southeast Spain

Crop rotations Na AIWS (mm) ARIS AWP (s m�3)

Autumn–spring sweet pepper 11 363 (30) 1.02 (29) 8.7 (40)

Sweet pepper–melon 14 502 (28) 1.07 (31) 11.4 (33)

Sweet pepper–green bean 4 489 (40) 1.01 (27) 12.9 (12)

Sweet pepper–watermelon 12 465 (25) 0.88 (30) 10.4 (29)

Cucumber–melon 17 486 (26) 1.45 (29) 12.1 (41)

Cucumber–watermelon 4 439 (41) 1.22 (24) 9.4 (40)

Green bean–green bean 12 365 (24) 1.16 (18) 15.3 (28)

a Number of measured greenhouses per crop rotation.

a g r i c u l t u r a l w a t e r m a n a g e m e n t 8 9 ( 2 0 0 7 ) 2 5 1 – 2 6 0 255

distribution uniformity (EU = 91%) and the low water salinity

(ECw = 1.05 dS m�1).

Values of annual irrigation water supply, AIWS (Table 3)

were generally higher than IWS values (Table 2) as most

greenhouse farmers tend to cultivate two crop cycles per

cropping season: average 6-year AIWS values for the main

crop rotations ranged between 363 mm for autumn–spring

sweet pepper and 502 mm for autumn–winter pepper and

spring melon (Table 3), while the mean AIWS for the main crop

rotations was 444 mm. Moreover, the CV of AIWS values

oscillated between 24% for the green bean–green bean rotation

and 41% for the cucumber–watermelon rotation (Table 3).

No information on IWS to horticultural crops by farmers is

available from other greenhouse irrigation areas with Medi-

terranean climate, since such studies have mainly been

conducted in open field irrigation areas (Lorite et al.,

2004a,b; Dechmi et al., 2003). Experimental data of IWS to a

greenhouse sweet pepper crop in Crete (Greece) are similar to

those measured in our study (Chartzoulakis and Drosos, 1997).

3.3. Irrigation performance indicators

3.3.1. Relative irrigation supply (RIS)Average 6-year RIS values for the various crop cycles evaluated

are summarized in Table 4. Mean RIS was close to 1 for

autumn–winter pepper, autumn–spring pepper and spring

cycles of melon, watermelon and green bean, and slightly

Table 4 – Mean values of relative irrigation supply percrop cycle (RIS) and during three crop periods along eachcycle: crop establishment (1), crop development (2), mid-to end of the season period (3)

RIS

Cycle Na 1 2 3

Autumn–winter pepper 0.95 (36) 12 2.78 1.27 0.91

Autumn–winter cucumber 1.63 (41) 15 3.53 1.48 1.20

Autumn–winter green bean 1.28 (24) 8 4.28 1.06 0.76

Spring melon 1.00 (39) 15 3.52 1.19 0.52

Spring watermelon 0.92 (33) 15 2.41 1.27 0.42

Spring green bean 1.03 (28) 10 4.25 1.80 0.60

Autumn–spring pepper 1.02 (29) 6 4.85 0.88 0.68

Campo de Dalıas (Almerıa), southeast Spain.a Number of selected greenhouses to evaluate RIS dynamics

throughout the crop cycle.

higher than 1 for autumn–winter green bean, indicating that,

on average, the irrigation supply matched the maximum

water requirements of these crops. By contrast, the mean RIS

of autumn–winter cucumber was 1.6, indicating that, on

average, the irrigation supply clearly exceeded the calculated

optima (Table 4). Cucumber is considered a very water

demanding crop by technical experts and local farmers,

who usually irrigate cucumber in excess (Vasco, 2003). On

the other hand, mean RIS values of greenhouse crops were

generally higher than those found for traditionally irrigated

outdoor crops in southern (Lorite et al., 2004b) and north-

eastern Spain (Dechmi et al., 2003). Because of the high value

of these greenhouse crops, water and nutrients are normally

applied at non-limiting rates.

The 6-year RIS values varied considerably among irrigators

(Table 4), with CV values ranging from 24% for autumn–winter

green bean to 41% for cucumber. These CV values were, in

general, similar to those found for traditionally irrigated

outdoor crops in southern Spain (Lorite et al., 2004b). Large

variability in RIS would indicate a substantial potential for

irrigation improvement, even if the mean RIS is reasonable. The

shape of the frequency distribution of RIS values was quasi-

normal for autumn–winter cycles of pepper and green bean,

autumn–spring pepper and spring cycles of melon, watermelon

and green bean, with most RIS values close to 1 (Fig. 2). For these

crops, the proportion of RIS values between 0.8 and 1.2 (which

may be considered around the optimum) varied from 30%

(autumn–winter pepper) to 55% (watermelon), the proportion of

RIS values less than 0.8 (greenhouses with theoretical irrigation

deficits) ranged from 10% (autumn–winter green bean) to 42%

(autumn–winter pepper), and the proportion of RIS values

higher than 1.2 (greenhouses with over-irrigation) oscillated

from 9% (spring watermelon) to 48% (autumn–winter green

bean). On the other hand, in autumn–winter cucumber crops

the RIS frequency distribution was biased towards values

higher than 1 (71% higher than 1.2 and 32% higher than 2) and

only 18% of them were between 0.8 and 1.2 (Fig. 2).

Analysis of the relative irrigation supply dynamics

throughout the crop cycles could provide deep insight into

greenhouse irrigation performance. In a representative num-

ber of greenhouses (Table 4), the relative irrigation supply was

determined during three consecutive crop periods: crop

establishment, crop development and mid to end cycle

(Doorenbos and Pruitt, 1977). On the whole, the relative

irrigation supply during crop establishment was very high

(between 2.41 and 4.85) for all crop cycles (Table 4). These

Fig. 2 – Relative irrigation supply (RIS) frequency distribution for the main greenhouse crop cycles on the west Almerıa coast.

N is the average number of greenhouses (%) for any given RIS interval.

a g r i c u l t u r a l w a t e r m a n a g e m e n t 8 9 ( 2 0 0 7 ) 2 5 1 – 2 6 0256

values were even higher when irrigation prior to sowing/

planting was considered, which, in turn, was greater before

autumn–winter cycles than between autumn and spring

cycles (data not shown). In the crop establishment period,

water requirements are usually low (�0.5 mm d�1), but farm-

ers often over-irrigate to improve both water availability when

the root system is small (Vasco, 2003) and microclimate

conditions (reducing air vapour pressure deficit by increasing

soil evaporation). Over-irrigation also ensured the leaching of

soil salts. This is a common irrigation practice that usually

results in high percolation water and nutrient losses in these

highly fertigated crops (Gallardo et al., 2006), but it can be

improved by increasing the irrigation frequency while

simultaneously reducing the water applied (Bonachela et al.,

2006). In the crop development period, the relative irrigation

supply was slightly higher than 1 in autumn–winter cycles of

pepper and green bean, and in melon and watermelon (1.19–

1.27), and clearly higher than 1 in cucumber and spring green

bean (1.48 and 1.80, respectively). Finally, during the mid to

end cycle period, the relative irrigation supply was lower than

1 for all of them (between 0.42 and 0.91), except in cucumber,

which was 1.20. Low RIS values at the end of crop cycles did not

usually imply the occurrence of crop water stress, especially

for short crop cycles, as: (i) the water buffering capacity of the

soil can offset moderate water deficits (Fernandez et al., 2005);

(ii) some rainfall may enter the greenhouse, especially in

winter periods and in flat roof greenhouses; (iii) some farmers

often deprive the crop of irrigation after the main harvest in

order to enhance maturing of the last fruits, especially in

spring cycles. In addition, the model used for predicting crop

water requirements does not consider some crop factors that

could reduce water consumption at the end of the cycle: (i)

crop ageing at the end of cycles, which normally reduces crop

transpiration (Allen et al., 1998), could be enhanced by extreme

microclimatic conditions, usually occurring at the end of

autumn–winter cycles and spring cycles in Mediterranean

greenhouses (Lorenzo, 1994; Orgaz et al., 2005); (ii) farmers

usually reduce the irrigation water supply at the end of melon

and watermelon crops in order to increase fruit quality

(Fernandez, 2000; Bonachela et al., 2006). On the other hand,

the high RIS values found early in the season, but decreasing

as the season progressed, have also been found in traditionally

irrigated outdoor crops in southern Spain (Lorite et al., 2004b)

and in other irrigation areas (Molden and Gates, 1990). Overall,

the mean RIS was close to 1 for most of the crop cycles studied.

However, the high CV values observed for RIS and the analysis

of the RIS dynamics throughout the cycles indicate that there

are greenhouse crops and crop cycle periods for which the IWS

clearly does not match the crop water requirements. Green-

house irrigation water use in the Almerıa coastal region can,

Table 5 – Mean values, averaged for six cropping seasons, and coefficients of variation (in parentheses) of cropproductivity (CP), irrigation water use efficiency based on marketable fruit production (IWUE) and water productivity (WP)

Crop cycle CP (kg m�2) WP (kg m�3) IWUE (s m�3)

Autumn–winter pepper 6.1 (23) 13.1 (50) 21.0 (40)

Autumn–winter cucumber 8.0 (27) 12.4 (59) 33.2 (52)

Autumn–winter green bean 2.3 (30) 15.9 (45) 15.3 (45)

Spring melon 3.8 (23) 10.1 (44) 22.8 (34)

Spring watermelon 6.0 (22) 7.8 (46) 35.6 (34)

Spring green bean 3.3 (43) 15.4 (53) 16.8 (31)

Autumn–spring pepper 5.9 (22) 8.7 (40) 16.9 (23)

Campo de Dalıas (Almerıa), southeast Spain.

a g r i c u l t u r a l w a t e r m a n a g e m e n t 8 9 ( 2 0 0 7 ) 2 5 1 – 2 6 0 257

therefore, be improved. We propose that farmers and

technical advisers employ simple models, already available

at http://www.fundacioncajamar.es to estimate irrigation

water requirements. These models could be used in combina-

tion with low-cost soil water sensors (Bonachela et al., 2006).

Finally, no information about RIS or ARIS indicators is

available from other Mediterranean greenhouse irrigation

areas.

3.3.2. Other irrigation performance indicatorsTable 5 shows values of crop productivity (CP), irrigation water

use efficiency based on marketable fruit production (IWUE)

and water productivity (WP). Values of CP ranged from

2.3 kg m�2 for autumn–winter green bean to 8.0 kg m�2 for

cucumber. These values are similar to mean CP values on the

Almerıa coast (Castilla and Hernandez, 2005) and in other

greenhouses areas in the Mediterranean basin, but much

lower than in high-technology glasshouses in Northern

European countries (Pardossi et al., 2004).

Water use efficiency is an important agronomic indicator

for areas with limited water resources (Sinclair et al., 1984;

Howell, 2001). Average 6-year IWUE values for greenhouse

horticultural crops varied from 15.3 kg m�3 for autumn–winter

green bean to 35.6 kg m�3 for spring watermelon (Table 5).

Similar or even higher values have been found for tomato

crops in climate-controlled soilless greenhouses (Pardossi

et al., 2004). Sweet pepper values were higher than those found

by Moller and Assouline (2007) in a screenhouse in Israel and

similar to those observed by Chartzoulakis and Drosos (1997)

in Crete (Greece). Compared to open field conditions, green-

houses markedly increase the IWUE for most greenhouse

horticultural crops [cucumber (Simsek et al., 2005; Ertek et al.,

2006); green bean (Sezen et al., 2005; Gencoglan et al., 2006);

melon (Fabeiro et al., 2002); pepper (Moreno et al., 2000; Sezen

et al., 2006); watermelon (Simsek et al., 2004)], except when

field crops receive a significant part of their water require-

ments via rainfall (Erdem and Yuksel, 2003; Orta et al., 2003).

This could be attributed to the lower IWS values applied to

greenhouse crops (Table 2) and, in some cases, to the higher

greenhouse crop productivity, compared to open field condi-

tions (Ertek et al., 2006; Sezen et al., 2005, 2006; Gencoglan

et al., 2006; Moreno et al., 2000). However, a comparison of

yield or IWUE data between greenhouse and open field

conditions should be considered with caution because crop

management and varieties are usually rather different.

Water productivity (WP) varied from 7.8 to 15.9 s m�3

(Table 5) and was highest for green bean crops due to their

low IWS and, especially, to the high prices of this commodity.

Spring watermelon and melon crops had the lowest WP values

due to their relatively lower market prices. The low IWS and,

above all, thehigh valueof the vegetable cropsgrown off-season

combine to yield WP values of 8–16 s m�3, which are generally

much higher than those found in other irrigation districts

around the world (Molden et al., 1998), including Mediterranean

areas (e.g. the mean WP of irrigated crops in the Andalucıa

region is about 1 s m�3, Junta de Andalucıa, 2003). Total water

cost for Almerıa greenhouse irrigators is high (0.1–0.3 s m�3,

Junta de Andalucıa, 2003) compared with other irrigated

Spanish areas (Lorite et al., 2004a; Dechmi et al., 2003). For

example in the Andalucıa region, including the Almerıa

greenhouse area, the total water price paid by irrigators is

about 0.04 s m�3 (Junta de Andalucıa, 2003). Water resources

are one of the main factors limiting further greenhouse

expansion on the Almerıa coast, but once farmers acquire a

water concession by the irrigation district, the water cost

becomes a normal production factor, as it only represents 4% of

total greenhouse production costs (Caja Rural de Almerıa, 1997).

This relatively low water cost could contribute to the over-

irrigation periods already mentioned for greenhouse crops

(Tables 4 and 5). Moreover, a substantial variability in the 6-year

WP values for the various crop cycles was found, with CV

ranging from 40% for autumn–spring pepper to 59% for

cucumber (Table 5). These high values could be, at least

partially, attributed to thehighintra-and inter-annualvariation

in non-subsidised vegetable prices (Puerto, 2001).

Finally, values of CP versus RIS were represented for each

crop cycle (Fig. 3). In the autumn–spring sweet pepper, CP was

significantly reduced when the RIS values decreased. In long

crop cycles, such as the autumn–spring sweet pepper,

irrigation supplies clearly lower than crop water requirements

during the mid- to end cycle period (Table 4) can affect CP

negatively. In the spring watermelon, CP was also significantly

reduced when the RIS decreased, which could be attributed to

very low RIS values during the mid to end cycle period

(Table 4). Other crop cycles studied did not show a relationship

between CP and RIS: several factors influencing crop produc-

tivity which are not considered in this study may mask these

relationships, such as microclimate conditions, pest and

diseases, crop types (California and Lamuyo pepper types;

Galia, Cantaloup and Piel de Sapo melon types; etc.), crop

varieties, fruit quality (pepper crop productivity is usually

lower when most fruits are harvested red than green), fruit

price (harvesting can be delayed until periods of higher prices),

crop management, etc.

Fig. 3 – Relationship between values of crop productivity (CP) and relative irrigation supply (RIS) for the main greenhouse

crop cycles on the west Almerıa coast.

a g r i c u l t u r a l w a t e r m a n a g e m e n t 8 9 ( 2 0 0 7 ) 2 5 1 – 2 6 0258

4. Conclusions

On the Southern Mediterranean Spanish coast, the mean IWS

value of greenhouse horticultural crops was 228 mm, whereas

the mean AIWS for the main crop rotations was 444 mm.

The mean irrigation supply matched the maximum water

requirements of most greenhouse vegetable crops, except for

autumn–winter cucumber where the mean irrigation supply

clearly exceeded the calculated optima. However, for most

crops the high CV values observed for RIS and the analysis of

the RIS dynamics throughout the cycles indicate that the

irrigation water use could be improved in a representative

group of greenhouse farmers. The mean irrigation supply

values clearly exceeded the maximum irrigation requirement

during early stages of most greenhouse horticultural crops.

Mean IWUE values for greenhouse horticultural crops

ranged from 15.3 to 35.6 kg m�3, while the WP varied from 7.8

to 15.9 s m�3 CP of autumn–spring sweet pepper and spring

watermelon was reduced when the RIS decreased.

Acknowledgments

This research was funded by a grant from Cajamar, Almerıa

University and FIAPA (Fundacion para la Investigacion Agraria

en la provincial de Almerıa). The authors would like to thank

Mr. J.C. Gazquez (Cajamar Foundation) for technical support

and the technical experts from the Campo de Dalıas irrigation

districts.

r e f e r e n c e s

Allen, R.G., Pereira, L.S., Raes, D., Martın, M., 1998. Cropevapotranspiration. In: Guidelines for Computing CropWater Requirements. FAO Irrigation and Drainage Paper 56,FAO, Rome.

Ayers, R.S., Westcot, D.W., 1985. Water Quality forAgriculture. FAO Irrigation and Drainage Paper 29, Rev. 1.FAO, Rome.

a g r i c u l t u r a l w a t e r m a n a g e m e n t 8 9 ( 2 0 0 7 ) 2 5 1 – 2 6 0 259

Bonachela, S., Gonzalez, A.M., Fernandez, M.D., 2006. Irrigationscheduling of plastic greenhouse vegetable crops based onhistorical weather data. Irrig. Sci. 25 (1), 53–62.

Caja Rural de Almerıa, 1997. Gestion del regadıo en el campo deDalıas: Las comunidades de regantes Sol y Arena y Sol-Poniente. Caja Rural de Almerıa, Almerıa, Spain.

Castilla, N., Hernandez, J., 2005. The plastic greenhouse industryin Spain. Chronica Hort. 45 (3), 15–20.

Chartzoulakis, K., Drosos, N., 1997. Water requirements ofgreenhouse grown pepper under drip irrigation. Acta Hort.449 (1), 75–80.

Dechmi, F., Playan, E., Faci, J.M., Tejero, M., 2003. Analysis of anirrigation district in northeastern Spain. I. Characterisationand water use assessment. Agric. Water Manage. 61, 75–92.

Doorenbos, J., Pruitt, W.O., 1977. Guidelines for Predicting CropWater Requirements. FAO Irrigation and Drainage Paper 24.FAO, Rome.

Enoch, H.Z., Enoch, Y., 1999. The history and geography of thegreenhouse. In: Stanhil, G., Enoch, H.Z. (Eds.), GreenhouseEcosystems. Ecosystems of the World 20. Elsevier,Amsterdam, pp. 1–15.

Erdem, Y., Yuksel, A.N., 2003. Yield response of watermelon toirrigation shortage. Sci. Hort. 98, 365–383.

Ertek, A., Sensoy, S., Gedik, I., Kucukyumuk, C., 2006. Irrigationscheduling based on pan evaporation values for cucumber(Cucumis sativus L.) grown under field conditions. Agric.Water Manage. 81, 159–172.

Fabeiro, C., Martın de Santa Olalla, F., de Juan, J.A., 2002.Production of muskmelon (Cucumis melo L.) under controlleddeficit irrigation in a semi-arid climate. Agric. WaterManage. 54, 93–105.

Faci, J.M., Bensaci, A., Slatni, A., Playan, E., 2000. A case study forirrigation modernisation. I. Characterisation of the districtand analysis of water delivery records. Agric. WaterManage. 42, 313–334.

Fernandez, M.D., 2000. Necesidades hıdricas y programacion deriegos en los cultivos hortıcolas en invernadero y sueloenarenado de Almerıa. Doctoral Thesis. UniversidadAlmerıa, Spain.

Fernandez, M.D., Gallardo, M., Bonachela, S., Orgaz, F.,Thompson, R.B., Fereres, E., 2005. Water use and productionof a greenhouse pepper crop under optimum and limitedwater supply. J. Hort. Sci. Biotechnol. 80 (1), 87–96.

Gallardo, M., Thompson, R.B., Fernandez, M.D., Lopez-Toral, J.,2006. Effect of applied N concentration in a fertigatedvegetable crop on soil solution nitrate and nitrate leachingloss. Acta Hort. 700, 221–224.

Gencogla˘n, C., Altunbey, H., Gencogla˘n, S., 2006. Response ofgreen bean (P. vulgaris L.) to subsurface drip irrigation andpartial rootzone-dryng irrigation. Agric. Water Manage. 84,274–280.

Gonzalez, A.M., 2003. Programas de riego para cultivoshortıcolas en invernaderos enarenados en Almerıa.Doctoral Thesis. Universidad Almerıa, Spain.

Howell, T.A., 2001. Enhancing water use efficiency in irrigatedagriculture. Agron. J. 93, 281–289.

Junta de Andalucıa, 2003. Inventario y caracterizacion de losregadıos de Andalucıa. Consejerıa de Agricultura y Pesca,ISBN: 84-8474-117-6 (CD Rom).

Lecina, S., Playan, E., Isidoro, D., Dechmi, F., Causape, J., Faci,J.M., 2005. Irrigation evaluation and simulation at theIrrigation District V of Bardenas (Spain). Agric. WaterManage. 73 (3), 223–245.

Lorenzo, P., 1994. Interceptacion de la luz, bioproductividad eintercambio gaseoso durante la ontogenia del cultivo depepino (Cucumis sativus L.) en invernadero. Doctoral Thesis.Universidad de Barcelona, Spain.

Lorite, I.J., Mateos, L., Fereres, E., 2004a. Evaluating irrigationperformance in a Mediterranean environment. I. Model and

general assessement of an irrigation scheme. Irrig. Sci. 23,77–84.

Lorite, I.J., Mateos, L., Fereres, E., 2004b. Evaluating irrigationperformance in a Mediterranean environment. II.Variability among crops and farmers. Irrig. Sci. 23, 85–92.

Malano, H., Burton, M., 2000. Guidelines for BenchmarkingPerformance in the Irrigation and Drainage Sector. IPTRIDSecretariat, FAO, Rome.

Merriam, J.L., Keller, J., 1978. Farm Irrigation System Evaluation.Utah State University Misc. Pub., 217 pp.

Molden, D.J., Gates, T.K., 1990. Performance measures forevaluation of irrigation-water-delivery systems. J. Irrig.Drainage Eng. 116 (6), 804–823.

Molden, D.J., Sakthivadivel, R., Perry, C.J., de Fraiture, C.,Kloezen, W.H., 1998. Indicators for comparing performanceof irrigated agricultural systems. Research Report 20.International Water Management Institute, Colombo, SriLanka.

Molden, D.J., Sakthivadivel, R., 1999. Water accounting to assessuse and productivity of water. Water Res. Dev. 15, 55–71.

Moller, M., Assouline, S., 2007. Effects of a shading screen onmicroclimate and crop water requirements. Irrig. Sci. 25,171–181.

Montero, J.I., Castilla, N., Gutierrez de Rave, E., Bretones, F.,1985. Climate under plastic in the Almerıa area. Acta Hort.170, 227–234.

Moreno, M.M. Caballero, M.J., Moreno, A., Ribas, F., 2000.Influencia de la dosis de riego sobre el rendimiento y laincidencia de necrosis apical en pimiento (Capsicum annuumL.). In: S.E.C.H. (Ed.), Acta Hort. 28, 159–166, Villaviciosa,Asturias, Spain, June 6–8, 2000.

Orgaz, F., Fernandez, M.D., Bonachela, S., Gallardo, M., Fereres,E., 2005. Evapotranspiration of horticultural cropsin an unheated plastic greenhouse. Agric. Water Manage.72, 81–96.

Orta, A.H., Erdem, Y., Erdem, T., 2003. Crop water stress indexfor watermelon. Sci. Hort. 98, 121–130.

Pardossi, A., Tognoni, F., Incrocci, L., 2004. Mediterraneangreenhouse technology. Chronica Hort. 44 (2), 28–34.

Perez Parra, J., Baeza, E., Montero, J.I., Bailey, B., 2004.Natural ventilation of parral greenhouses. Biosyst. Eng. 87,355–366.

Puerto, H., 2001. Respuesta productiva de un cultivo de pepinoholandes bajo invernadero a la calefaccion por aire caliente.Doctoral Thesis. Universidad de Almerıa, Spain.

Pulido-Bosch, A., Bensi, S., Molina, L., Vallejos, A., Calaforra,J.M., Pulido-Leboeuf, P., 2000. Nitrates as indicator of aquiferinter-connection. Application to the Campo de Dalıas (SE-Spain). Environ. Geol. 39, 791–799.

Sanjuan, J.F., 2004. Estudio Multitemporal sobre la Evolucion dela Superficie Invernada en la Provincia de Almerıa porTerminos Municipales desde 1984 hasta 2004 medianteTeledeteccion de Imagenes Thematic Mapper de losSatelites Landsat V y VII. FIAPA, Almerıa, Spain.

Seckler, D., 1996. The new era of water resources management:from ‘dry’ to ‘wet’ water savings. Research Report 1.International Irrigation Management Institute, Colombo, SriLanka.

Sezen, S.M., Yazar, A., Cambolat, M., Eker, S., Celikel, G., 2005.Effect of drip irrigation management on yield and qualityof field grown green beans. Agric. Water Manage. 71 (3),243–255.

Sezen, S.M., Yazar, A., Eker, S., 2006. Effect of drip irrigationmanagement on yield and quality of field grown bellpepper. Agric. Water Manage. 81, 115–131.

Simsek, M., Kacira, M., Tonkaz, T., 2004. The effects of differentirrigation regimes on watermelon (Citrullus lanatus (Thumb))yield and yield components under semi-arid conditions.Aust. J. Agric. Res. 55 (11), 1149–1157.

a g r i c u l t u r a l w a t e r m a n a g e m e n t 8 9 ( 2 0 0 7 ) 2 5 1 – 2 6 0260

Simsek, M., Tonkaz, T., Kacira, M., Comlecioglu, N., Dogan, Z.,2005. The effects of different irrigation regimes oncucumber (Cucumis sativus L.) yield and yield characteristicsunder open field conditions. Agric. Water Manage. 73 (3),173–191.

Sinclair, T.R., Tanner, C.B., Bennett, J.M., 1984. Water-useefficiency in crop production. BioScience 34, 36–40.

Vasco, R., 2003. El cultivo de pepino en Almerıa. In: Cajamar(Caja Rural Intermediterranea), Instituto de Estudios deCajamar (Eds.), Tecnicas de produccion de frutas yhortalizas en los cultivos protegidos, vol. 2. Cajamar,Almerıa, Spain, pp. 692–722.

Wittwer, S.H., Castilla, N., 1995. Protected cultivation ofhorticultural crops worldwide. HortTechnology 3, 6–19.