Introduction

The application of growth bioregulators is a management practice that is performed in some temperate, tropical and subtropical fruit trees with different purposes, such as regulating vegetative growth, stimulating flowering and/ or increasing fruit binding. This is the case of Paclobutrazol “PBZ”, which retards vegetative growth and decreases plant vigor due to the reduction of gibberellin levels (^{Ramírez et al., 2005}; ^{Ramírez et al., 2008}; ^{Kishore et al., 2015}). ^{Rademacher (2000)} and ^{Kishore et al. (2015)} mention that plant growth retardants such as PBZ and calcium prohexadione slow the cell division and elongation by physiologically regulating plant height without causing deformation in leaves or stems. However, the complex effect of PBZ with other hormones can lead to different responses, highlighting its use as a promoter of flowering, early flowering, reducing alternation, decreasing vegetative sprouting and increasing yield (^{Cárdenas and Rojas, 2003}; ^{Ramírez et al., 2003}; ^{Pérez et al., 2011}; ^{Kishore, 2015}; ^{Pérez et al., 2016}).

^{Pérez et al. (2011)}, reported the effects of PBZ on mango manila on flowering, harvest season and fruit yield, with emphasis on the improvement in flowering rate, crop yield and fruit yield increase; mentioned that with the use of PBZ they achieved a 100% increase in yield with respect to the control.

^{Cárdenas and Rojas (2003)} mentioned that the application of paclobutrazol restricted vegetative growth and stimulated the floral development of ‘Tommy Atkins’ mango, with an earlier bloom and greater number of fruits per inflorescence, although it did not affect the percentage of retained fruits.

Regarding the effect of PBZ and other similar products such as calcium Prohexadione on plant growth, ^{Pérez et al. (2016)} reported a reduction of the vegetative growth of ‘Ataulfo’ mango trees from 18 to 41%, with the PBZ being the one that caused the lowest reduction percentage with respect to Prohexadione of calcium.

In horticultural species, the effect of paclobutrazol has also been evaluated. ^{Partida et al. (2007)} reported the application of paclobutrazol on root growth and aerial biomass in red pepper and eggplant seedlings. They mention that the PBZ increased the root biomass and aerial part of both species in comparison to the control.

Preliminary studies on guava cultivation in a high density plot report the use of paclobutrazol in established high density plants, where it is mentioned that the PBZ in a production cycle significantly increased yield and number of fruits and a reduction was observed in the length of the sprouts, the leaf area and the distance between knots, as well as an increase in chlorophyll content measured as SPAD units (^{Padilla et al., 2014a}; ^{Padilla et al., 2014b}). Subsequently, the ex post effect of PBZ was reported in the following cycle without application of the product, whose results indicated that the yield and number of fruits of all plants showed a strong reduction, attributed to a possible alternating effect and not to an effect of the PBZ applied in the previous cycle (^{Padilla et al., 2015}).

Regarding planting density, in the main guava producing areas in Mexico, traditional plantations were carried out with densities of 204 to 266 trees ha^-1. Under this traditional 7 m × 7 m planting system, guava produces on average in Calvillo, Aguascalientes around 15 t ha^-1 (^{SIAP-SAGARPA, 2017}); However, when selected genotypes are established at high densities (3×3 m), yields were higher than 30 t ha^-1 (^{Padilla et al., 2007}). Newly established guava gardens in the Calvillo-Cañones region, which includes the states of Aguascalientes and Zacatecas, as well as Michoacán, have been established at higher planting densities in order to increase their productivity.

In India, a number of studies have been conducted in Guava to evaluate high density systems, ranging from 3 333 to 10 000 plants ha^-1, reporting significant yield increases per plant at densities of 2 × 1.5 m (8.43 kg plant^-1), compared to 5.89 kg plant^-1 in the density of 1 × 1 m; however, yield per hectare was significantly higher in planting densities of 1 × 1 m (58.93 t ha^-1), while in the density of 2 × 1.5 m the production was 28.85 t ha^-1 (^{Pal y Lal, 2015}). ^{Brar (2010)} reported the effect of PBZ and etephon on vegetative growth of guava established at planting densities from 333 to 833 plants ha^-1, and noted that the PBZ significantly reduced plant height compared to the control (3.49 m vs 3.79 m, respectively), as well as the cup volume (39.17 m3 vs 53.81 m3), so that the use of growth regulators such as PBZ and increase planting density can contribute to increase the productivity of the guava crop considering economic aspects and fruit quality.

^{Kishore et al. (2015)} mention that when PBZ is applied at optimal doses, soil and fruit residues have not been detected in quantities above the quantifiable level (0.01 ppm), suggesting the use of PBZ in adequate doses to obtain maximum benefits without undesirable effects.

In addition to the application of PBZ, the use of genotypes with greater potential for fruit yield and high density planting systems can contribute to maintaining a balanced growth and allowing an adequate balance of vegetative and reproductive development. Therefore, the objective of the present study was to evaluate the effect of the application of “PBZ” on the growth dynamics of sprouts of the year and on the fruit yield of eight guava genotypes established in a high density system.

Materials and methods

The study was carried out at the “Los Cañones” Experimental Site, belonging to the National Institute of Forestry, Agriculture and Livestock Research (INIFAP) and located in the municipality of Huanusco, Zacatecas (latitude 21° 44.7’ north, longitude 102° 58.0’ west, 1 580 meters above sea level).

The predominant climate in the study region is BS1 hw(w), which corresponds to a semi-warm semi-dry climate, according to ^{García (1981)} classification. The annual average temperature is 20.1 °C, the maximum annual average is 30.4 °C and the minimum is 9.8 °C. These temperatures are similar to those prevailing in the main guava producing areas of the country, where the average annual temperature fluctuates between 20 and 23 °C (^{Padilla et al., 1999}). The annual accumulation of heat units (Tb= 9 °C) in the area is an average of 4 060, which are considered sufficient to satisfy the thermal requirement of guava (^{Ruiz et al., 1992}; ^{Padilla et al., 2012}). The average annual rainfall is 516 mm, of which more than 80% occur from June to September (Figure 1).

Figure 1 Mean, minimum and maximum temperature averages and accumulated precipitation in decadal periods from 2006 to 2015 in the Experimental Site Los Cañones del INIFAP. Huanusco, Zacatecas. 

The physical -chemical characteristics of the soil at the experimental site are: sandy-loam texture, alkaline pH of 8.2, with 1.16% of organic matter and NPK of 28.9, 0.65 and 2 345 ppm, respectively. This indicates a level of soil fertility suitable for N and K, although for P it is considered very low. On the other hand, considering the soil texture and values of field capacity (20.6%) and permanent wilting point (10.9%), the soil presents a low moisture retention capacity, therefore the application of frequent watering.

The lot has a system of micro-sprinkler irrigation, by which irrigations are applied weekly during the eight months of the crop cycle. Fertilization was provided through the irrigation system, where the dose applied during the cycle was 90-90 -90 for N- P2O5-K2O, respectively, covering the stages of sprouting, flowering, mooring and fruit development. Weed control was done manually and the pests and diseases according to the recommendations of the Experimental Field Pavilion for the cultivation of guava in the Calvillo-Cañones region.

It is important to mention that as part of the guava management in the study region, the annual production cycle is initiated after the “calmeo”, which is a common practice during which the application of irrigation for 2 to 4 months is suspended. Induces the plant to a state of quiescence due to the imposed water stress. This practice is performed at the end of the harvest of the previous cycle and is used to schedule the harvest of the new cycle, obtaining this one between 6 to 8 months after the application of the first irrigation. Another reason for calmeo is to avoid frost damage in trees during the winter period (^{Perales et al., 2002}).

For the application of the treatments the commercial product Austar® (Paclobutrazol 25%) was used during two cycles of production, where the application of the PBZ was done alternately in 2013 and 2015. In 2013, three doses were evaluated: 0, 1 and 2 ml tree^-1, whereas in 2015 only two doses were applied: 0 and 1 ml tree^-1, of the commercial product. The treatments were applied to the soil at the base of the stem in one liter of water, for which a trench 10 cm deep was dug around the base of the stem where the solution was applied and immediately afterwards the trench was again covered. In the case of control trees without PBZ, the same procedure was performed, applying only water. The application of the PBZ was carried out in the last week of March in both years and three to four days of pruning to start the annual cycle of production (in the cycle 2014 PBZ was not applied

The guava genotypes evaluated were: three varieties (Calvillo S-XXI, Hidrozac and Caxcana) and five selections (S-12, S-20, S-45, S-46 and S-47) belonging to the Germplasm Bank of INIFAP. The three varieties used correspond to the group of varieties released and registered by the INIFAP-Pavilion Experimental Field (Padilla et al., 2010). Ten plants were established for each genotype under field conditions in the spring of 2008, with material propagated by aerial the previous cycle of the “mother” plants of each variety or selection of the INIFAP germplasm bank, located at the “Los Cañones” Experimental Site. The plants are established in a high density system with 2 m of separation between lines and 1 m between plants (5 000 plants ha^-1) during the evaluation, the plants were five and seven years old, and the characteristics of the eight genotypes evaluated are shown in Table 1.

The variables evaluated included the following aspects of plant growth and fruit production:

Growth dynamics of sprout. The dynamics of sprout growth were determined for a period of 102 days (19 April to 30 July), data recording started three weeks after PBZ application, for which five sprout plant^-1 in a plant of each treatment, recording weekly the length of the sprouts marked with a flexible tape measure with graduation to decimals of centimeter (0.0 cm). The length of internodes was measured 140 days after pruning. In addition, the daily growth rate (TCD) in mm day^-1 was estimated between every two dates of the growth dynamics record of the outbreaks according to the following equation.

Where: TCD= daily growth rate (mm day^-1); LB_t2= sprout length at time t₂ (mm); LBt1= sprout length at time t1 (mm) t2 - t₁= time interval between t1 to t₂ (days).

Fruit yield. During the harvest, which was carried out when the fruits presented a green-yellow to yellow color corresponding to stages 3 or 4 of the ripening scale for guava proposed by ^{Padilla et al. (2002)}, the total number of fruits tree^-1, fruit production (kg tree^-1) and average weight per fruit (g fruit^-1) were recorded. The harvesting period started from the second week of October and ended at the end of November, between 200 and 250 days after the application of the PBZ.

The information from the variables recorded was analyzed based on a randomized block design with split plot arrangement, where the large plot were the genotypes and the small plot were the PBZ doses and were considered three replicates in 2013 and four in 2015 for each treatment. The experimental unit consisted of a plant. Data were analyzed using the statistical package SAS version 8 (^{SAS Institute, 1999}) and when significance was detected between treatments, the DMS test was applied with a significance of 0.05.

Results and discussion

Sprout growth

In the first cycle of application of the PBZ (2013), the length of the sprouts and internodes was significantly reduced in the plants treated with PBZ in comparison to the control. The average length of sprouts in plants without PBZ was 68.9 cm whereas with 1 and 2 ml L^-1 of PBZ the sprout length was 52.7 and 40.3 cm, respectively. This means that the outbreaks had a growth reduction of 23.5 and 41.5% compared to the control (Figure 2).

Figure 2 Length of sprouts in 2013 (above) and 2015 (below) in guava plants treated with PBZ. Each point represents the average of 40 sprouts (five plant^-1 sprouts per eight genotypes). The bar represents the typical error. 

The TCD showed a decreasing pattern in all treatments, being higher at the beginning of sprouting (5 to 8 mm day^-1) and lower (1 to 3 mm day^-1) as the bud increased its length. The average TCD of sprouts was 6.10 mm day^-1 in plants without PBZ, whereas with 1 and 2 ml of PBZ, the growth rate was 3.96 and 2.81 mm day^-1, respectively (Figure 3). The average length of the internodes for the eight genotypes of guava was 4.13 cm with a range of 3.83 cm in Calvillo S-XXI to 4.60 cm in S-12. The PBZ caused reductions of internodes between 22.8 and 30.8% with the doses of 1 and 2 ml L^-1 of PBZ, respectively (Table 2).

Figure 3 Daily growth rate in 2013 (above) and 2015 (below) of guava plants treated with PBZ. Each point represents the average of 40 sprouts (five plant^-1 sprouts in eight genotypes). The bar represents the typical standard error. 

Table 2 Length of internodes in sprouts of eight guava genotypes treated with different doses of PBZ. 

Factores de evaluación	Longitud de entrenudos (cm)
	2013	2015
Genotipos
Calvillo S-XXI	3.83	4.2
S-12	4.6	4.7
S-20	4.03	4.5
S-45	3.8	4.1
S-46	4.13	4.6
S-47	4.17	4.5
Caxcana	3.97	4.4
Hidrozac	4.47	5
DMS0.05	0.51	ns
Dosis de PBZ (ml L-1)
0	5.03	4.9
1	3.88 (22.8)	4.10 (16.3)
2	3.48 (30.8)	-
DMS0.05	0.31	0.28

ns= no significativo. El número entre paréntesis indica el porcentaje de reducción, respecto al testigo.

In the cycle 2015 of PBZ application only showed significant effects on the length of internodes, but not on the length of the sprouts, however, a reduction of sprout size was observed in plants treated with PBZ (52.8 cm) to untreated plants (62.4 cm), which represents 15% less compared to the control (Figure 2). The TCD showed a pattern similar to that observed in 2013, which is attributed to the competition of the fruits, which are already in development at this stage. The average TCD of the sprouts was 4.53 mm day^-1 in the plants without PBZ, whereas with 1 ml L^-1 of PBZ, it was 3.90 mm day^-1 (Figure 3).

In the 2015 cycle, the average length of the internodes in the eight genotypes was 4.5 cm, and ranged from 4.1 cm in the S-45 to 5 cm in the variety Hidrozac. The treatment of PBZ showed an effect similar to that observed in 2013, although the reduction of the internodes was slightly lower, which was 16.3% (4.9 vs. 4.1 cm with 0 and 1 ml L^-1 of PBZ, respectively) (Table 2). Similar effects on the reduction of internodes of two apple varieties when applied prohexadione-Ca were reported by (^{Ramírez et al ., 2003}), mentioning reductions of up to 46% in the “Royal Gala” variety.

The percentages observed in this work of the reduction of the growth of sprouts of guava plants treated with PBZ were similar to those found by ^{Pérez et al. (2016)} who reported a reduction between 18 and 41% of the growth of “Ataulfo” mango sprouts with calcium prohexadione (P-Ca) and 6% with PBZ.

The average TCD observed in plants without PBZ was slightly higher than reported by ^{Damián et al. (2004)} who found a growth rate between 2.8 and 3.2 mm day^-1 in guava plants established in a dry tropic climate, although these values are similar to TCD when PBZ was applied.

^{Ramírez et al. (2005)} reported a significant reduction in the height of tomato plants when applied Prohexadione calcium that acts as a growth retardant similar to PBZ, however, the treated plants showed higher number of clusters and fruits, weight and firmness of fruit and production per plant.

Conclusions

The application of PBZ in guava plants in high density, showed in both years a reduction in the length of sprouts and internodes, which was reflected in a lower growth rate day^-1.

The annual alternating application of PBZ in guayabo high density showed inconsistency in both production cycles, in the first application, the PBZ increased yield and fruit number, while in the second opposite response was obtained.

As for the genotypes, the Calvillo S-XXI variety and the S-45 and S-12 selections stand out with the highest average fruit yield in both production cycles. It is important to consider the interaction of the genotype with the application of the PBZ, trying to optimize a production system in high density.