Introduction

Soilless systems allow an adequate control of growth and plant development helping to obtain high yields. However, these production systems require frequent irrigation and high rates of fertilization and when free-draining (open system) is used caused ground and surface pollution of water sources (^{Van Os, 1999}). The management of closed cultivation systems offer excellent advantages in terms of limiting the problem of water and nutrients loss, it also allows a more efficient production and environmentally friendly compared to open crop systems (^{Van Os, 1999}; ^{Ahmed et al., 2000}; ^{Siddiqui et al., 1998}; ^{Rouphael et al., 2004}; ^{Rouphael and Colla, 2005}).

Reducing of fertilizer and water requirements for crop production under greenhouse has become increasingly important for producers, since many are facing higher costs of water and fertilizer, decreased water quality and government regulations to protect surface and groundwater (^{Van Os, 1999}, ^{Uva et al., 2001}). A promising alternative to be more efficient in the production of crops of importance is the adoption of ubirrigation systems with recirculating nutrient solution, also referred to as zero runoff subirrigation (^{Uva et al., 2001}; ^{Santamaria et al., 2003}; ^{Rouphael et al., 2006}). This system works by allowing water to move from a reservoir where nutrient solution (SN) is stored to a tray application within which are containers, keeping SN for a certain time to allowit to move through the culture medium by capillary action (Bouchaaba et al., 2015). After irrigation is complete, the amount of SN that is not absorbed by the medium, is returned back to the storage tank to be reused in subsequent irrigations (^{van Os, 1999}; ^{Incrocci et al., 2006}; ^{Pinto et al., 2008}), for which is necessary to perform periodic adjustments to water volume, pH and nutrient concentration, the latter being assessed through EC measurement (^{Cox, 2001}; ^{Incrocci et al., 2006}).

Subirrigation systems offers many advantages, such as a lower nutrients and water requiremnts, provides nutrients in a uniform manner, prevents leaf wetness (disease prevention), irrigation uniformity, less substrate compaction, more uniform crops, improved productivity; reduced discharge of nutrients into surrounding ecosystems and reduce production costs (^{Cox, 2001}; ^{Santamaria et al., 2003}; ^{Rouphael and Colla, 2005}; ^{Rouphael et al., 2008}; ^{Montesano et al., 2010}). These benefits generate savings in labor, material inputs and output losses (^{Purvis et al., 2000}; ^{Santamaria et al., 2003}). Also, subirrigation systems can facilitate SN management as it maintains stable parameters of the same, since the elements that are not absorbed by the plant accumulate in the upper part of the substrate instead of accumulating in SN as it would in an open irrigation system (^{Reed, 1996}; ^{Kent and Reed, 1996}; ^{Morvant et al., 1997}; ^{Santamaria et al., 2003}; ^{Rouphael and Colla, 2005}; ^{Rouphael et al., 2006}; ^{Montesano et al., 2010}).

However, the tendency of salt accumulation on the upper part of the growth medium represents a disadvantage for subirrigation systems as it can result in reduction of crop growth, especially in long-term crops and in dry and hot environmental conditions (^{Kent and Reed, 1996}; ^{Reed, 1996}; ^{Morvant et al., 1997}; ^{Cox, 2001}; Bouchaaba et al., 2015). Salts accumulating on the upper part of the growth medium can occur if SN is too concentrated, because the growth medium does not leach during production (^{Martinetti et al., 2008}). Therefore, fertilizers concentration in subirrigation systems should be lower than in surface irrigation systems (^{Broschat Klock-Moore, 1999}; ^{Cox, 2001}; ^{Mak and Yeh, 2001}; ^{Yeh et al., 2004}; ^{Martinetti et al., 2008}).

Several advantages of subirrigation system have been reported for ornamental plants under greenhouse production, however, it has been paid less attention to this technique for vegetable production (^{James and Van Iersel, 2001}; ^{Santamaria et al., 2003}; ^{Serio et al., 2004}). It is necessary to verify the validity of these systems for vegetable production under greenhouse, as these are characterized by a long crop cycle, have a high growth rate and high demand for water and nutrients (^{Santamaria et al., 2003}; ^{Rouphael and Colla, 2005}), besides studying the suitability of different cultivars to this irrigation system as consequence of its tolerance to salinity (^{Martinetti et al., 2008}). The present study aimed to determine some subirrigation system requirements for pepper production and its effect on growth, yield and some chemical characteristics of the substrate.

Results

In general, the container of 25 L obtained a higher VAR than 13 L, although CH was higher with container of 13 L (Table 1). Greater VAR and CH were observed when water lamina was 15 cm and immersion time was longer (Table 1). Similarly, at higher height on the substrate layer there was a tendency to decrease both VAR and CH (Table 1).

Table 1 Behavior of retained water volume and moisture content in each layer of the container according to irrigation lamina, immersion time and volume from the container used. 

Medias con la misma letra en cada columna son iguales de acuerdo con la prueba de comparación múltiple de Tukey con p≤ 0.05.

The interaction between the factors under study (Table 1) suggests that VAR in the strata was greater by using immersion times of 30 min in both laminas, observing this trend in both types of container (Figure 1); however, this difference was more pronounced when a lamina of 15 cm (Figure 1) was used. When the solution lamina was 15 cm and this was maintained for 20 min, VAR was similar to that obtained when the lamina was 10 cm maintained for 30 min in both types of container. VAR was decreasing as strata height was greater in the container, retaining higher volume in the highest strata when lamina of the solution was 15 cm and this was maintained for 30 min (Figure 1).

Figure 1 Retained water volume in the layers of the containers 13 and 25 L, two irrigation lamina (10 and 15 cm) and three immersion times (10, 20 and 30 min) in the solution. The bars represent the standard error of the mean. 

CH was higher in the lowest strata of the container, but in relative terms, this was decreasing as immersion time increased from 10 to 30 min, regardless of lamina used (Figure 2). In the strata 7-14 cm CH was lower with immersion times of 30 min in both lamina of the container 13 L; however, the container 25 L, CH of these strata was greater with a longer immersion time (Figure 2). In strata 14-21 cm, CH was greater as immersion time increased in both containers, regardless of the lamina used (Figure 2). In the upper layer of the container 13 L, CH was higher with an immersion time of 30 min, on the contrary, with container 25 L, CH from the upper layer decreased with this immersion time (Figure 2).

Figure 2 Moisture content in the layers from the containers with volume of 13 and 25 L, two irrigation lamina (10 and 15 cm) and three immersion times (10, 20 and 30 min) in the solution. 

Compared with surface irrigation, fruit number was not affected by subirrigation treatments (Table 2); however, fruit yield and harvest index were higher in plants subjected to subirrigation with a lamina of 15 cm for 20 min, similar to those obtained by plants with surface irrigation (Table 2).

Table 2 Effect of irrigation system (subirrigation and surface irrigation) in fruit production of pepper plants (Capsicum annuum L.) grown in container of 13 L. 

Medias con la misma letra en cada columna son iguales de acuerdo con la prueba de comparación múltiple de Tukey con p≤ 0.05.

The dry weight of leaves was negatively affected by subirrigation, obtaining higher leaf biomass in plants treated with surface irrigation (Table 3), while dry weight of stem, root and total dry weight were higher in plants subjected to a subirrigation lamina of 15 cm for 30 min (Table 3). These changes in the distribution of dry weight were reflected in a change of the relationship between shoot/root, as compared to surface irrigation plants, subirrigated plants showed relatively greater root development than in shoot (Table 3).

Table 3 Effect of irrigation system (subirrigation and surface irrigation) on dry weight of pepper (Capsicum annuum L.) grown in container of 13 L. 

Medias con la misma letra en cada columna son iguales de acuerdo con la prueba de comparación múltiple de Tukey con p≤ 0.05.

Upon completion of the study, the average pH in the substrate increased when plants were subirrigated with a lamina of 10 cm for 30 min, while EC was higher when using a lamina of 15 cm for 30 min (Table 4). The concentration of Ca2⁺ and K⁺ in the substrate was higher when plants were managed with surface irrigation; however, the concentration of NO₃^- was higher when used a subirrigation lamina of 15 cm with an immersion time of 20 min (Table 4). The pH was more acidic in the strata 1421 cm while EC and Ca²⁺ concentration were higher in the upper strata of the container (Table 4). The concentration of K⁺ and NO₃^- tended to be higher in the lower strata of the container (Table 4).

Table 4 Behavior of pH, electrical conductivity (EC) and concentration of calcium (Ca²⁺), potassium (K⁺) and nitrate (NO₃^-) from the substrate in function of the strata and irrigation system, irrigation lamina and immersion time. 

Medias con la misma letra en cada columna son iguales de acuerdo con la prueba de comparación múltiple de Tukey con p≤ 0.05.

The interaction between the factors under study (Table 4) suggests that irrigation systems used during crop development, as well as lamina and immersion time evaluated in subirrigation systems showed different effect on the chemical properties of the substrate at the end study (Figure 3). The pH in the stratas increased in plants subirrigated with 15 cm lamina maintained for 20 min and in plants with surface irrigation, being higher in the 21-28 cm strata while this was reduced when the lamina was 10 and 15 cm maintained for 20 and 30 min, respectively (Figure 3). EC increased as strata height from the container raised, being higher in strata 21-28 cm; however, substrate from plants subirrigated with lamina of 15 cm for 30 min, the EC was higher in all stratas in the remaining subirrigation treatments or in plants with surface irrigation (Figure 3). Ca²⁺ concentration tended to increase in the upper strata, being higher with surface irrigation (Figure 3). NO₃^- concentration decreased as strata height increases in the container, being lower in 21-28 cm strata in subirrigation treatments; in contrast with surface irrigation, the concentration of NO₃^- tended to increase with strata height, being higher in 21-28 cm strata (Figure 3). K⁺ concentration tended to decrease with height of the stratas in the container in plants with subirrigation; however, in plants with superficial irrigation, K⁺ increased in the upper strata (Figure 3).

Figure 3 Effect of the subirrigation lamina and immersion time on pH, electrical conductivity (EC), and calcium (Ca²⁺), potassium (K⁺) and nitrate (NO₃^-) in the stratas of the container. The bars represent the standard error of the mean. 

Some of the chemical properties of the substrate at the end of the study were correlated with fruit yield, as this tended to increase quadratically when average concentration of NO₃^-, K⁺, and Ca²⁺ increased, while pH levels of 5.26 were associated with increased production (Figure 4).

Figure 4 Relationship between average concentration of nitrate (NO₃^-), potassium (K⁺), calcium (Ca²⁺) and pH from substrate with fruit yield in pepper (Capsicum annuum L.) grown in container of 13 L. The bars represent the standard error of the mean. 

Discussion

Proportionally, substrate moisture was higher in the container of 13 L than in 25 L, however, this was even greater when subirrigation lamina was 15 cm maintained for 30 min, confirming that mentioned by ^{NeSmith and Duval (1998)}; ^{Vence (2008)} in the sense that percentage of retained volume of water by the substrate depends on height, diameter, volume and shape of the container. With this management subirrigación obtained a more uniform moisture profile in container distribution, although the highest strata always had the lowest CH, an important factor in early stages of the crop as this is in where the roots are established after transplantation.

Lower CH suggests that capillary movement of water through peat requires more time and greater irrigation lamina to reach the highest stratas, however, the irrigation lamina should not be sustained for a longer time due to prolonged anoxic conditions since can affect root plants. ^{Reed (1996)} mentions that in subirrigation systems SN should not be maintained for a long period (over 45 min), otherwise it may damage the roots by waterlogging, noting that ideally 10 to 15 minutes and that flood depth only has to be 2 to 2.5 cm or, flood about 20% to 25% of the height of the container. Similar conditions in terms of lamina and immersion time have been used to perform research with subirrigation; e.g. ^{Whitcher et al. (2005)}; ^{Richards and Reed (2004)} used a lamina of 2 cm for 10 min to produce impatient New Guinea (Impatiens hawkeri Bull.) in a container with 4.6 L of substrate while ^{Incrocci et al. (2006)} used a lamina of 2-2.2 cm for 15 min in tomato (Solanum lycopersicon L.) in a container with 3.2 L. The lamina and immersion time in SN used in the aforementioned studies differ from those used in this study, mainly because the required container for the production of vegetables necessarily must be of higher volume, as ideal lamina and immersion time for subirrigation will depend on the species with which works, water needs, composition and substrate volume, and container dimensions.

VAR was directly proportional to substrate volume contained in each container as the 25 L container retained more water than 13 L, however, the container of 13 L; the retained volume increases when a lamina of 15 cm sustained for 30 min was used. This effect was made in the container of 25 L only in strata from the central part (7-14 and 14-21 cm). These observations suggest that in subirrigation the diameter of the container affects the capillary movement of water, and as the diameter of the container increases water movement to upper strata is slower. The shape and size of the container determines the three-dimensional distribution of the substrate, which may considerably influence plant yield as the dimensions define air-filled porosity and water retention capacity by the same and this will depend largely on its physical properties (^{Da Silva et al., 1993}; ^{Gizas and Savvas, 2007}).

In general, the vegetative growth of the plants was higher when subirrigated with a lamina of 15 cm for 30 min, or when received surface irrigation, since the dry weight of the plants was higher. This response was related to EC since the increase of this chemical property of the substrate obtained an increase total dry weight. Opposite results to those obtained in this study were reported by ^{Kent and Reed (1996)} in impatient New Guinea and peace lilies (Spathiphyllum Schott) and by ^{Rouphael and Colla (2005)}, who found that with drip irrigation biomass production in zucchini was 44% higher than with subirrigation as a result of the lower EC. ^{Incrocci et al. (2006)} reported that high EC in the upper part of the substrate caused no salinity stress in tomato under subirrigation, arguing that roots grew mostly in the lower strata of the substrate; the authors reported no differences between plants treated with subirrigation and drip irrigation in terms of plant growth, which coincides with the results of this study.

^{Santamaría et al. (2003)}; ^{Scoggins (2005)} mention that drip irrigation and subirrigation systems determine a different stratification of salts in the culture medium, which concentrate in the lower part with the first method and in the top with the second and depending on the species, these levels may or may not pose a problem. The results of this study do not fully agree with the above because although an increase was detected in salt concentration in the upper strata of the substrate in subirrigated plants, this also was detected in plants with drip irrigation; however, the impact of surface irrigation on salinity of the substrate if detected in the central stratas. Bouchaaba et al. (2015) state that the excessive salinity produced in the substrate can have dramatic effects on root growth of plant that are particularly sensitive to salt stress by the presence of possible osmotic stress due to higher salinity reached in subirrigated substrates.

Although the vegetative growth of organs was promoted in subirrigated plants with a lamina of 15 cm maintained for 30 min, this was not reflected in increased fruit production, since yield was higher in plants subirrigated with a lamina of 15 cm maintained for 20 min as well as in plants with surface irrigation. This may be because pH and EC of the 0-7, 7-14 and 14-21 cm stratas as well as other chemical properties thereof, maintained at levels closer to the optimum than other treatments from subirrigation and surface irrigation. Fruit yield was associated with an increase in average concentration of NO₃^-, K⁺ and Ca²⁺ in the substrate, suggesting that the irrigation method used can dramatically impact plant productivity with subirrigation systems. ^{Rouphael et al. (2008)} mention the importance of maintaining a favorable EC in lower strata of the substrate to maintain optimum crop yield due to the presence of greater proportion from the root system at this depth.

The results reported by ^{Santamaría et al. (2003)}); ^{Scholberg and Locascio (1999)} differ from those obtained in this research, reporting that with drip irrigation tomato production was greater than with subirrigation as a result of EC in subirrigation. ^{Martinetti et al. (2008)} mention that the difference regarding yield of subirrigation systems and drip irrigation is due to with subirrigation EC concentrates on the upper part, while with drip irrigation EC is lower and is distributed uniformly throughout the volume of the substrate, which was confirmed in this study.

The fact that plants obtained higher yield did not matched to a higher production of dry matter which reflected in harvest index, suggesting that subirrigated plants with a lamina of 15 cm maintained for 20 min as well as surface irrigation, a more favorable biomass distribution towards fruit production was established. Although substrate from subirrigated plants with lamina of 15 cm maintained for 30 min was obtained with higher CH, this was not favorable for plants because it promoted higher growth of vegetative parts instead of promoting fruit production. This also suggests that in the case of pepper, subirrigation laminas should not be sustained for a period over 20 minutes, because the prolonged anoxia conditions may be affecting growth.

The initial pH of the substrate was 6.1, however, at the end of the experiment this was acidified in function of the strata and the applied treatment, which may be due to the extrusion of hydrogen ion (H⁺) when the plant absorbs cations (^{Voogt, 1995}). In treatments where the highest yield (lamina 15 cm maintained for 20 min and the surface irrigation) was obtained, pH tended to increase as strata height rises, being higher in the upper strata, suggesting that in the lower part of the container, the most acidified and where a greater amount of root accumulates, nutrient uptake was more intense. An opposite behavior was presented by subirrigating with laminas of 10 cm for 20 min and 15 cm for 30 min and pH tended to decrease in the upper strata, suggesting that the highest activity of nutrient uptake was carried in the upper part of the substrate. ^{Martinetti et al. (2008)} reported that with subirrigation pH was more acidic in upper layers of the substrate, while with drip irrigation pH remained stable in the different strata of the container.

EC from substrate at the beginning of the study was 0.6 dS m^-1, but at the end of the experiment was 1.02-1.2 dS m^-1 in the lower strata to 1.56-2.23 dS m^-1 in the upper strata, having a rising behavior to the upper strata in all treatments. A similar pattern of salt accumulation with subirrigation systems has been reported in numerous studies (^{Cox, 2001}; ^{Zheng et al., 2004}). ^{Kang and van Iersel (2001)} mentions that salt accumulation in the culture medium depends on salt concentration applied with SN, the irrigation system, and the demand for environmental evaporation. By implementing a lamina of 15 cm for 30 min EC was higher in each of the stratas, which may be due to with this lamina and immersion obtained increased volume of SN in substrate profile, which increased mineral accumulation applied in SN. The highest EC in the upper strata with subirrigation is due to the movement of water inside the substrate is enhanced by the capillary force, mass flow of nutrients, selective uptake of minerals by the root (^{Rouphael and Colla, 2005}; ^{Reed, 1996}; ^{Incrocci et al., 2006}; ^{Rouphael et al, 2006}) and by evaporation from the substrate surface (^{Rouphael et al, 2008}); in contrast, EC tended to be lower in each strata when the surface irrigation was used. Bouchaaba et al. (2015) ^{Incrocci et al. (2006)} associated the increase of EC of the growth medium to increased concentrations of Na ⁺ and Cl^- in the upper strata of the substrate while ^{Incrocci et al. (2006)} found that K ⁺ does not affect the increase of ECin subirrigation systems.

The Ca²⁺ tended to increase in the upper strata of both irrigation systems, being of 358 ppm with drip irrigation and 142-197 ppm with subirrigation, suggesting that Ca²⁺ movement to the upper strata was affected by subirrigation. K ⁺ and NO₃^- tended to decrease in the upper strata in subirrigated plants, contrary to what happened with drip irrigation. In other studies with subirrigation has been reported the accumulation of K ⁺ (^{Haley and Reed, 2004}; ^{Richards and Reed, 2004}; ^{Zheng et al., 2004}; ^{Martinetti et al, 2008}; ^{Montesano et al, 2010}), NO₃^- (^{Zheng et al., 2004}; ^{Martinetti et al, 2008}; ^{Montesano et al, 2010}), H₂PO₄^-, Mg²⁺, Ca² and Na ⁺ (^{Zheng et al, 2004}; ^{Martinetti et al, 2008}) in the upper strata of the container.

Conclusions

Pepper production under subirrigation systems is feasible as yields are similar to those obtained when grown with surface irrigation. To take advantage of subirrigation in pepper, the SN lamina must be 15 cm and this has to be maintained for a period of 20 min to achieve favorable water retention for this species.