1. Introduction
In the current climate change scenario, the availability of water for human activities is expected to decrease [
1]. Of these activities, agriculture is the one that consumes the most water, with irrigated agriculture accounting for 70% of freshwater extraction, and this percentage can reach 90% in some regions. Of the total water used, about 20% comes from groundwater sources (renewable or not), and this proportion is increasing rapidly, especially in dry areas [
2]. Given this situation, with limited water resources, the use of drip irrigation systems is increasing throughout the world. This irrigation system can provide water to the plant frequently and directly in the root zone of the plant [
3], being able to achieve a very high irrigation efficiency. Not only is the irrigation system important, irrigation management must also be efficient to minimize the possible environmental impact derived from it, as well as promote the sustainable use of resources [
4,
5]. In cultivation systems where drip irrigation systems are used, the use of fertigation for the supply of nutrients has become widespread, so in these systems soil is not irrigated, it is fertigated.
Horticultural crops have different fertigation needs depending on factors such as: Type of crop, stage of development and production system. Despite of the fact that greenhouse horticultural crops are very efficient in the use of water and fertilizers compared to other production systems, it is necessary to increase efficiency, for which individual fertigation strategies must be designed for each crop and each situation.
The water and nutritional needs of greenhouse horticultural crops have been extensively studied. The works that determine the water needs are mainly focused on the predictive calculation of crop evapotranspiration (ETc) with average climatic data or real-time data [
6,
7,
8]. However, for the same crop, the different varieties show different vegetative development (some are more vegetative than others). Furthermore, the growing conditions of the crop can affect the vegetative development of the crop. For this reason, the most recent research has been directed towards the dynamic determination of the crop coefficient (K
c) in situ, using photo digitization systems that monitor plant growth [
9], and towards the use of humidity sensors of the soil that allow irrigation on demand [
4].
Regarding fertigation, most of the research focuses on determining standard nutrient solutions for each crop and stage [
10], as well as determining the extraction and absorption curves of the different crops [
11,
12,
13,
14]. As in irrigation, the different vegetative development generates, for the same crop, depending on the variety and growing conditions, different nutritional needs. Hence, the most recent investigations focus on the dynamic determination of these needs in situ through: Use of optical sensors to determine the nitrogen content in crops [
15,
16], use of systems for monitoring the level of nutrients in the soil solution [
13,
17,
18] and development and adaptation of different simulation models of nutrient requirements [
19,
20].
The cultivation system has a great influence on the consumption of water and fertilizers [
20,
21,
22]. Within the greenhouse cultivation system there are several factors that determine and modify the consumption of water and fertilizers, since they modify the environmental conditions in which they develop, such as: Greenhouse structure, ventilation and type of soil. [
23,
24,
25,
26]. Not only do the structural factors have influence, but so do the cultivation techniques such as pruning, phytosanitary treatments, since they modify the index of leaf area and the biomass of the cultivation.
Since the late 1980s, great advances in electronics and information technologies have led to significant progress in the development, availability, and application of sensors for use in irrigation scheduling and automation. Electrotensiometers are soil matric potential (SMP) sensors for a continuous control of water application by a computer. The automation of irrigation using electrotensiometers can be a viable option at the farm level that, among other advantages, offers the possibility of watering according to the individual characteristics of greenhouses and crops and can provide a more precise adjustment of the irrigation frequency to the crop needs at every moment of the cycle, minimizing losses due to drainage to deeper layers. In addition, limiting drainage, also is preventing nutrient leaching that can cause contamination of aquifers. However, for automatic irrigation control based on a value of the soil matric potential to be effective, it is essential to establish an adequate value from which irrigation begins. This threshold value depends on the crop species, its development stage, evaporative conditions and soil texture [
27].
Many of the studies carried out for a specific species show a wide range of threshold values of soil matric potential, which suggests, as pointed out by Thompson et al. [
28] the influence of site-specific factors. Therefore, it is very important to establish the appropriate value of threshold values of soil matric potential for each crop and specific development condition that optimize production and efficiency in the use of water and nutrients [
4,
29,
30].
The pepper is one of the most important horticultural crops (
Capsicum anuum L.) for greenhouse production. Specifically, in Spain, its production is concentrated on the Mediterranean coast of Andalusia, highlighting Almería as the main growing area with more than 11,000 ha and an annual production of 845,595 tons [
31].
The objective of this work was to determine the influence of the fertigation activation command with electrotensiometers and the characteristics of the greenhouse on the water and nutrients productivity in a pepper crop.
2. Materials and Methods
2.1. Experimental Site
The trial was carried out at the IFAPA La Mojonera Center, Almería, in two multitunnel greenhouses with a semi-elliptic curved roof of the same surface (900 m
2) and orientation (east–west), with a passive climate, with a metal structure, plastic polyethylene cover and windows lateral and zenithal. In both greenhouses, the cultivation system was soil with addition of layer of sand about 5 cm, characteristic of intensive horticultural crops in the Southeast of the Peninsula [
32].
2.2. Experimental Design and Cropping Systems
The experimental design used balanced incomplete block 2 by 2 factorial design with eight replicates (four per greenhouse), resulting in the combination of the two factors studied in the four treatments tested. Each greenhouse was divided into 4 blocks and in each one of them the treatments were placed randomly distributed, resulting in 16 experimental plots per greenhouse (
Figure 1). A single plot measured 12 m by 4 m.
The factors studied were: SMP and greenhouse (G).
Two levels of SMP have been established, SMP
−10: Automatic activation of irrigation through an electronic tensiometer when the level of SMP was −10 kPa, and SMP
−20: Automatic activation of irrigation through an electronic tensiometer when the level of SMP was −20 kPa. The timing of irrigation was constant throughout the cycle (being 30 min) and the same for two levels of SMP. The frequency of irrigation depended on the value of the SMP established in each treatment. In a previous experiment, we established three levels of soil matric potential −10 kPa, −20 kPa and −30 kPa, as well as a fourth treatment that received the irrigation calculated with the irrigation scheduling based on crop evapotranspiration. Under the development conditions, the soil matric tension threshold of −20 kPa was the one that had shown the best results considering the water and nutrient productivity criteria, but −10 kPa was the highest fruit yield [
33]. For this reason, in this experiment we have studied these two levels.
The total volumes of irrigation applied were different in each treatment, since irrigation was managed by the threshold level of soil matric potential established for each treatment. What remained constant in all treatments was the irrigation activation period, which was from 10:00 a.m. to 6:00 p.m. in winter and from 8:00 a.m. to 8:00 p.m. in spring. A 2 h pause was also established after each irrigation to ensure the correct response for the acquisition of tensiometric data.
Two different greenhouses have been studied: G
1 and G
2. The differences between the greenhouses were greenhouse height, size and the type of lateral and zenith windows, and the properties of the soil. Greenhouse G
1 was more height than G
2 (4.2 m of the gutter height and the total ridge height was 6 m) and more height lateral windows (1.7 m), and had zenith windows at the top of a quarter arch (
Figure 2) and soil with a clay loam texture (
Table 1). Greenhouse G
2 had zenith windows at the top of a half arch with an opening in the gutter (
Figure 2), 3.6 m of the height to gutter and the total height to ridge was 5.4 m, height lateral windows of 1.25 m and soil with a clay texture (
Table 1).
To determine the physical-chemical parameters in the soil twenty random sample points of each replication were selected to be taken and the samples of soil were mixed to achieve a representative soil sample per replication. The samples of soil were dried in a forced air oven (MEMMERT Model 800) at 50 °C for 48 h and then were gridded and screened (2 mm).
Texture of soils (clay, silt and sand percentages) was determined by Bouyoucos-hydrometer analysis [
34]. pH and electrical conductivity (EC) were determined in the saturated extract by pHmeter (model MicropH 2002 Crison, Hach Large Spain S.L.U., Barcelona, Spain) and conductivity meter (model GLP31 Crison, Hach Large Spain S.L.U., Barcelona, Spain), respectively. Sodium adsorption ratio (SAR) was determined in the water extract from saturated soil paste. Organic matter was determined using the Walkley–Black method [
35].
The interpretation of the soil-water characteristic curves (SWCC) for each soil are show in
Table 2. SWCC describes the amount of water retained in a soil (expressed as volume water content, θv) under equilibrium at a given matric potential. The model used to predict such relationship was Van Genuchten [
36]. An SWCC is an important hydraulic property, related to size and connectedness of pore spaces, hence strongly affected by soil texture and structure, and by other constituents, including organic matter. The soil of the G
1 greenhouse has a higher θv than the soil of the G
2 greenhouse for the same SMP level (
Table 2).
Irrigation was applied by drip irrigation. Pressure-compensating and self-closing drippers were employed (PCJ Dripper—Netafim®). The flow rate of the emitters was 3 L h−1 and two emitters per square meter were installed (spaced at 0.5 m in the drip line and 1 m between lines).
The concentration of the nutrient solution applied was the same for all treatments and remained constant throughout the cultivation cycle, being in mM: 12 of NO3−, 1.5 of H2PO4−, 1.5 of SO42−, 6.0 of K+, 5.0 of Ca2+ and 2.0 of Mg2+.
Pepper plants (Capsicum anuum L. type Lamuyo cv. Mazo) were transplanted on 5 September 2018 with a plant density of 2 plants m−2 and the cycle ended on 23 April 2019.
2.3. Measurements
Climatic parameters in greenhouses: Air temperature and air relative humidity inside the two greenhouses were measured. The measurements were recorded every 30 min during 24 h a day. Daily mean vapor pressure deficit (VPD) was calculated for each greenhouse.
Soil matric potential was measured in the most representative area of the plant roots. For this, 32 electrotensiometers (Irrometer Co., Inc., Riverside, CA, USA) were installed, 8 per treatment. The electrotensiometers were placed 15 cm deep and 15 cm from the base of the plant. Soil matric tension measurements were automatically recorded by a Red Himarcan® System control device.
Soil drainage volume was determined for each treatment. For this, each greenhouse was equipped with 16 drainage lysimeters (one for treatment and replication) of 1 m2 of surface, installed at a depth of 50 cm, discounting the upper layer of sand. The volume of water drained was measured daily throughout the experiment.
Irrigation water and nutrients applied were measured by installing three volumetric water meters, model M120 (Elster, Iberconta S.A., Gipuzkoa, Spain), one for treatment, also was corroborated with the registration number of irrigations performed and the volume applied in each irrigation. The concentration of nutrients (NO3−, P, K+, Ca2+ and Mg2+) in the applied nutrient solution was analyzed in the laboratory weekly. The samples were collected at the outlet by the dropper, placing a carafe that collected the volume of fertigation applied weekly in each treatment.
Yield of the pepper crop was evaluated by manually harvesting the red fruits. The production of 48 plants was controlled per replication and treatment, resulting in a total of 384 plants per treatment. The harvest period began on 9 January 2019 and ended on 23 April 2019, with a total of 5 harvests. The weight and number of marketable and unmarketable fruits were counted. The marketable fruits were separated by size. The average weight of the fruits by size was determined.
Vegetative growth of pepper plants was determined by measuring the dry biomass and the percentage of dry matter of the aerial part of the plant. Crop harvest index (HI, g g−1) was calculated as the ratio between generative dry biomass and total shoot dry biomass. For this, whole plants (leave, stem and fruit), excluding roots, were collected from each experimental plot (six plants per repetition) at the end of the experiment and 20 fruits per experimental plot were also randomly selected in each harvest. The fresh samples were weighed and dried at 70 °C to constant weight and the dry weight was determined.
Productivity of water and nutrients was calculated. There are numerous authors who have defined the productivity of water or the efficiency of water use [
37,
38,
39,
40,
41,
42]. We define the productivity the following way:
Water productivity (WP); Y/W.
Nitrogen productivity (NP); Y/Nc.
Phosphorus productivity (PP); Y/P.
Potassium productivity (KP); Y/K.
Magnesium productivity (MgP); Y/Mg.
where:
Y is fruit marketable yield (kg m−2).
W is water applied (m3 m−2).
N is nitrogen applied (kg m−2).
P is phosphorus applied (kg m−2).
P is potassium applied (kg m−2).
Ca is calcium applied (kg m−2).
Mg is magnesium applied (kg m−2).
2.4. Statistical Analysis
An analysis of variance (ANOVA) was performed for a 2 by 2 factorial balanced incomplete block design with eight repetitions. To identify the significant factors (SMP and G) and the interactions between the factors, a multifactorial ANOVA was performed. When the ANOVA was significant, an least significant difference LSD test (p ≤ 0.05) was performed to identify statistically significant differences between the means. To obtain a normal distribution, the percentage data were transformed with an inverse sign √. The statistical software used was Statgraphics 18.
5. Conclusions
The results obtained in this work showed the great importance of establishing an optimum level of SMP threshold value to automate fertigation with electrotensiometers in greenhouse pepper crops. Under the development conditions of the experiment, the activation threshold of −10 kPa increased fruit production as well as the biomass and conserved water and nutrients productivity with respect to the activation threshold of −20 kPa in both greenhouses.
The greenhouse factor also had a significant influence on fruit production, the crop in greenhouse G1 obtained higher yields and vegetative development, also requiring a greater volume of fertigation than the crop developed in the greenhouse G2 for the same activation threshold. This better response was associated with a more favorable average matric potential recorded in the soil of this greenhouse, since the matric potential threshold after irrigation was closer to zero. Therefore, it was shown that not only the SMP threshold used by fertigation activation had an influence, but also the lowest SMP threshold obtained after fertigation, confirming that it is necessary to establish depth irrigation that achieve a SMP closer to zero to achieve increased fruit yield and crop growth. Therefore, it is essential to know the soil retention curve to adapt the irrigation endowment and reach a potential close to zero without producing losses due to deep filtration (drainage).
The automation of fertigation with electrotensiometers allowed applying the volume of fertigation demanded by the crop, which was different according to the soil conditions of each greenhouse, and was also different depending on the threshold value of the soil matric potential established for activation and also allowed to eliminate water losses by drainage and therefore the leaching of nutrients.