Characteristics of the soil

Five trees were selected randomly in each orchard, and a soil sample was taken made up of four sub-samples from the drip zone of the tree, at 0-30 cm of depth, which is the zone with highest amounts of young roots (^{Salazar-García and Cortés-Flores, 1986}). The analyses were performed in a laboratory accredited by the North American Proficiency Testing (NAPT) program (a program by the Soil Science Society of America). The variables were texture, pH (1:2 water) (^{McLean, 1982}), organic matter (OM) by the Walkley and Black method (^{Nelson and Sommers, 1982}), inorganic-N (^{Dahnke, 1990}), P-Bray (^{Bray and Kurtz, 1945}), K, Ca, Mg and Na with ammonium acetate (^{Doll and Lucas, 1973}), Fe, Zn, Cu, Mn by DTPA method (^{Lindsay and Norvell, 1978}), and B was determined by the hot water and Azomethine-H method (^{Bingham, 1982}). The variables P and B were determined in a Genesys 20 spectrophotometer (Thermo Scientific, Madison, USA); an atomic absorption spectrophotometer (Thermo Series S, Shangai, China) was used for the other nutrients.

Phenological variables

Twenty shoots per tree were tagged in 10 trees per orchard, at the beginning of the vegetative growth flushes of winter and summer 2014 and 2015. From each tree and shoot marked, an apical bud or flower shoot was collected at biweekly or weekly intervals as anthesis was approached. The buds and flower shoots were fixed in FAA (formaldehyde:acetic acid:ethanol, 5:5:90, v:v:v) and then they were introduced into a vacuum bell (Nalgene 8040317, Nalgen Company) at 30 KPa for 5 h. Later they were classified with a stereoscopic microscope (Zeiss Stereomikroskop Mod. Stemi 2000-C, Carl Zeiss, Göttingen, Germany), with the visual scale suggested by ^{Salazar-García et al. (1998)} that covers from S-1 (vegetative bud) to S-11 (anthesis).

In each orchard, another group of 10 trees was selected, with a history of annual production of at least 50 kg per tree, and they were used in all the assessments except those of flower development. In February and August 2014 and 2015, 30 shoots were labeled on each tree at the beginning of each vegetative flush (winter and summer, respectively). The types of growth produced (floral, vegetative or inactive shoots) by the tagged shoots were quantified at the end of each flowering period, in October and February 2014, and in November 2015 and March 2016.

The root samples were taken each month from February 2014 to February 2016, alternating odd and even trees, as well as north and south orientation of the trees. The young roots, of light brown color, in a soil volume of 40 ( 40 ( 40 cm in the drip zone of the tree, were extracted, washed and dehydrated in a forced air oven (Binder Mod. ED 240, Binder Corp., Tutllingen, Germany) at 70 °C for 72 h, to obtain their dry weight.

The fruit drop was quantified each year for fruits from the summer and winter flowering periods. In anthesis, five floral shoots were marked in each cardinal point from the medium part of the canopy from every 10 trees selected. For fruit from the summer flowering, the number of fruit were counted per flower shoot, two (November) and six (March) months after anthesis. In fruit from the winter flowering, this was done two (April) and five (July) months after anthesis.

Five fruit were labeled in each tree marked, from each flowering period, when they were 2 cm long. Each month the length and diameter were measured with a digital Vernier scale (Mitutoyo, Mod. CD-6’’ CSX, Mitutoyo Corp., Kawasaki, Japan) from October (fruit from the summer flowering) and April (fruit from the winter flowering).

Climate variables

The data of rainfall were obtained from the National Meteorological Service (^{Servicio Meteorológico Nacional, SMN, 2016}). The environmental and soil (at 30 cm of depth) temperatures were recorded in each orchard with battery-operated automatized thermometers (HOBO H8, Onset Computer, Witzprod, Englewood Cliffs, NJ, USA).

Experimental design an statistical analysis

The experimental design was random blocks with a 2 ( 2 factorial arrangement: orchards (Colorín 1 and Ocote Cuate 2), and vegetative flushes (summer and winter), with 10 replicates (trees); and the experimental unit was 30 shoots per tree. The ANOVA was performed with repeated measurements for each orchard and vegetative flush. Prior to their analysis, the values expressed as percentage were transformed by the arcsine x+0.5 (^{Steel and Torrie, 1984}). The means were compared with the Tukey test (p≤0.05) and the ^{SAS V. 9.3 software was used (2011)}.

Soil fertility

The characteristics of soil fertility were similar between orchards. The OM, Ca and Mg ranged from moderately low to low and in the Colorín 1 orchard the low levels of Mn, B and Zn were evident (Table 1). The low content of OM in addition to the insufficient levels of Mn, B and Zn are characteristic of soils of volcanic origin (^{Salazar-García et al., 2016}), suggesting the need to incorporate OM and nutrients in moderately low to low concentrations to improve the soil fertility. The pH was neutral in both orchards (Table 1). In other avocado production regions, the soils present similar or more acidic pH (^{Aguilera-Montañez and Salazar-García, 1991}; ^{Salazar-García and Lazcano-Ferrat, 1999}). However, avocado can grow in pH of 4.8 or higher than 6.5 (^{Salazar-García, 2002}).

Type of growth

The production of flower shoots had different behavior between orchards. In Colorín 1, 33.1 % was found and in Ocote Cuate 2, 19.1 % (Tables 2 and 3). The decrease in environmental temperature is the factor that promotes floral development in avocado (^{Salazar-García et al., 1999}, ^2006a, ²⁰¹³). The difference in production of floral shoots between orchards is explained by the lower temperatures in Colorín 1 (annual average 20.5 °C) compared to Ocote Cuate 2 (22.5 °C) (Figures 1a and 1b), which favored a greater exposure of the El Colorín 1 orchard to cooler temperatures. The vegetative flush of origin (winter or summer) did not affect the type of growth produced (vegetative, floral or inactive) (Tables 2 and 3).

Figure 1 Environmental (A and C) and soil (B and D) temperature recorded in the Colorín 1 and Ocote Cuate 2 orchards, as well as rainfall (D). FVI: winter vegetative flush; FVV: summer vegetative flush. 

Vegetative growth

The phenological record began with the budbreak (February 2014) of shoots from the winter vegetative flush. Out of 100 % of winter shoots marked, 25 % flowered in the summer (September 2014) and 6 % flowered the following winter (February 2015). Of the winter shoots, 55 % produced summer shoots in September 2014, and of these, 27 % flowered in winter (February 2015). The winter and summer shoots produced an important number of vegetative shoots (between 50 and 60 %) (Figure 2). In contrast with ‘Méndez’, in ‘Hass’ the winter flush does not produce flowering in the summer; however, it can happen if the environmental conditions are favorable (^{Rocha-Arroyo et al., 2011a}).

Figure 2 Origin and destination of the vegetative growth flushes (V.F.) of the ‘Méndez’ avocado. 

In ‘Hass’ the winter vegetative flush (known as spring flush in other regions) is the most important for the winter-spring flowering of the following year, and it is documented in Mexico (^{Cossio-Vargas et al., 2008}; ^{Rocha-Arroyo et al., 2011b}) and other countries (^{Mena-Volker, 2004}; ^{Robinson et al., 2002}; ^{Salazar-García and Lovatt, 1998}; ^{Thorp et al., 1995}; ^{Whiley et al., 1990}).

Flower development

The duration of flower development (from vegetative shoot to anthesis) showed slight variation between years. For 2014, 218 and 180 d went by for the winter and summer shoots, respectively. The 1^st or summer flowering presented anthesis in September, and the 2^nd or winter flowering, in February (Figure 3). In 2015, the winter shoots completed flower development in 225 d, producing the 1^st flowering between the first week of September and the first of October. In the case of summer shoots, the flower development required 193 d, with anthesis on the first week of March 2016 (2^nd flowering).

Figure 3 Main phenological stages of ‘Méndez’ avocado in the southern region of the state of Jalisco. 1^st flowering: summer flowering; 2^nd flowering: winter flowering. 

In the two years of study, the summer shoots required, in average, 35 d less than the winter ones to complete the flower development. This difference was lower than the 120 and 180 d mentioned for both flushes of ‘Hass’ in Nayarit and Michoacán, respectively (^{Cossio-Vargas et al., 2008}; ^{Rocha-Arroyo et al., 2011a}). The lesser time required for the summer ‘Méndez’ shoots to complete their flower development could be because when they emerged (August-September) the environmental temperature had already begun its descent (21 °C in average), stimulating floral development. The effect of the temperature decrease in developing ‘Hass’ shoots was mentioned by ^{Rocha-Arroyo et al. (2010)} and ^{Salazar-García et al. (2013)}.

Fruit drop

The fruit dropped in the so-called “second drop”, “physiological drop” or “June drop” (northern hemisphere) presented diameters between 5 and 6 cm. For fruits from the first flowering period (summer), the drop was 61.8 % and it happened in March. The fruit drop from the second flowering period (winter) was in July (55.4 %). The magnitude of the physiological drop recorded in our study is similar to the one described for ‘Hass’ in other production regions: 45 % in Nayarit, Mexico (^{Cossio-Vargas et al., 2008}) and 54 % in South Africa (^{Toerien, 1979}). The age of the dropped fruit from the summer and winter flowering periods was 22 and 26 weeks after anthesis, respectively. The fruit drop from the second flowering was shortly before the summer vegetative flush and for those from the first flowering, it was after the winter vegetative flush (Figure 3).

^{Cossio-Vargas et al. (2008)} and ^{Lovatt (1990)} mention the coincidence, and ^{Wolstenholme and Whiley (1989)} the lack of coincidence between the fruit drop and the presence of vegetative growth, but there is no conclusive information about the causes that originate the second fruit drop (^{Dixon et al., 2006}). A hypothesis mentions the hormonal imbalance as responsible for this fruit drop, and highlights abscisic acid (ABA) as primary cause (^{Adato and Gazit, 1977}). In order to decrease the intensity of this drop in ‘Hass’ avocado, leaf spraying with 2,4-D or aminoethoxyvinylglycine were evaluated because two months before the beginning of the fruit fall, and in the second year of study, it decreased 18 and 17.7 %, respectively; in addition, it increased the production of fruit of large calibers (170 to ≥ 266 g fruit^-1) in 32 and 23.6 %, respectively, compared to the control (^{Salazar-García et al., 2006b}). ^{Garner and Lovatt (2016)} confirmed that ABA is involved in the second fruit drop in ‘Hass’ avocado in California.

Fruit growth

The ‘Méndez’ fruit that result from the summer and winter flowering periods required 10 and eight months, respectively, to reach physiological maturity (22.7 % d.m.) (Figure 3). For fruit from the winter flowering, there are similar results for ‘Hass’ (21.5 % d.m.) in Nayarit and Michoacán (^{Cossio-Vargas et al., 2008}; ^{Rocha-Arroyo et al., 2011a}). The fruit from the summer flowering reached physiological maturity in July, with an average diameter of 6.5 cm; and those from the winter flowering between September and October, with an average diameter of 6.4 cm. The fruits derived from the summer flowering required two more months to reach physiological maturity, which can be attributed to the fact that when they began their growth (September-October), the air (21.8 °C) and soil (20.8 °C) temperatures had begun their descent, which decreased the rhythm of tree metabolism.

Root production

Root production did not stop during the year and there were two periods of higher intensity (Figure 3). The most important flush was in the rainy months of the summer (June-August), prior to the summer vegetative flush. The second flush of root growth was less intense, from February to March, and coincided with the winter flowering and its corresponding vegetative flush. The highest production of roots in the summer coincided with what is described for ‘Hass’ cultivated without irrigation in Nayarit (^{Cossio-Vargas et al., 2008}), and with and without irrigation for the same cultivar in Michoacán (^{Bárcenas-Ortega et al., 2007}). In Israel, ^{Kalmar and Lahav (1976)} found in ‘Ettinger’ and ‘Fuerte’ a positive association between the amount of water applied and the root production, but in Michoacán, ^{Bárcenas-Ortega et al. (2007)} documented the opposite in ‘Hass’.

According to ^{Bernstein et al. (2004)}, in avocado, root growth can be more restricted by the salinity than by the presence of vegetative shoots. In our study, in both orchards the CE of the soil varied from 0.26 to 0.33 dS m^-1, which is why it was not a limiting factor for the development of roots and did not affect the fruit drop. In addition, the lowest production of roots happened when the fruit was at its maximum growth (Figure 3).

Rainfall and air and soil temperatures

In both orchards, the annual rainfall was 728 mm, concentrated between June and September (Figure 1). In the Colorín 1 orchard, the maximum and minimum average temperatures of the warmest (April and May) and coldest (January and February) months were 32, 31 and 6.4 and 6.0 °C, respectively (Figure 1A). In the Ocote Cuate 2 orchard the temperatures of the warmest (May and June) and coldest (March) months were 30.8 and 30.6 and 15.0, respectively (Figure 1C).

The highest soil temperature (23.8 °C) was in May and the lowest (14.5 °C) in March in Colorín 1 (Figure 1B). For Ocote Cuate 2 the maximum soil temperature was in May (25.2 °C) and the minimum in January (15.9 °C) and February (16.0 °C) (Figure 1D). In both orchards, the maximum soil temperatures remained between 21 and 27 °C, mentioned by ^{Yusof et al. (1969)} as adequate for the growth of avocado roots. Likewise, the minimum temperature was higher than 13 °C, pointed out by ^{Whiley et al. (1990)} as the lower limit from which avocado root growth stops.