Location of the study area

The present research was developed in the agricultural cycles autumn-winter 2015-2016 and spring-summer 2016, two irrigation schedules and two planting densities, for each cycle. Experimental lots were established in the Valley of Fuerte Experimental Field-INIFAP, with geographic location at 25° 46’ 492’’ north latitude and 108° 48’ 181’’ west latitude with an altitude of 4 masl (http://www.inegi. org.mx). Its climate is generally humid warm, and has an average annual temperature of 25.9 °C. There is a minimum annual temperature of 18 °C and an annual maximum of 33.9 °C, the hottest season being from May to October. In the reference period, the average rainfall is 383 millimeters per year, being the rainiest months from July to October (http://www.weatherbase.com).

Experiment management

Genotypes evaluated. The hybrids used were: Maravilla, Chinatan, HB-8, and 2B-5, plus four commercial controls: Olga, Eva, K-93 and Zoya (Table 1), from China and Argentina , which were planted and harvested manually in the Valley of Fuerte Experimental Field (Guasave, Sinaloa), during the autumn-winter 2015-2016 and spring-summer 2016 cycles.

Experimental design

A factorial design with four repetitions was used, with four levels of observation: genotypes, two agricultural cycles, two planting densities and two irrigation schedules.

Variables evaluated

The plant characteristics considered were plant height (AP, cm), days to flowering (DF, days), end of flowering (FDF, days), days to maturity (DM, days), bunch length first generation (LRG1), bunch weight per generation (PRG, g), weight of a thousand seeds (PMS, g) and grain yield (REND, kg ha^-1), as established by ^{Goytia et al. (2011)}.

Data analysis and experimental model

An analysis of variance of a factorial design was performed with an arrangement in divided plots using the SASv9.4 program (2015), and the Tukey test (p≤ 0.05) for the comparison of means between the genotypes.

Combined factor analysis

Significance was found in hybrids, irrigations and dates in seven evaluated plant variables, the interaction hybrids by dates of reading was significant for beginning of flowering, end of flowering, days to maturity, weight of bunch in first generation, bunch length in first generation and yield (Table 2).

According to the analysis of the information in relation to hybrids, density, irrigations and dates, interactions H*F, D*F or R*F.

Differences were detected in seven evaluated plant variables, irrigation was only significant for yield and for dates it was days to flowering, end of flowering and physiological maturity of the evaluated hybrids, the hybrid interaction by date was significant for six variables, of which two of them have a significant impact on yield, such as bunch length in 1st generation and bunch weight in the 1st generation (LRG and PRG (Table 2).

Variation of plant morphological characteristics between hybrids

In terms of days to flowering of the hybrids, 2-B5, Chinatan and HB-8 were earlier with days to flowering ranging from 64 to 69, days respectively. In fact, the difference of days from the beginning of flowering to harvest was 76 days for both Maravilla and HB-8, and 88 for 2B-5. For Chinatan it was 84 days, which indicates that HB-8 has a higher grain filling rate (TLLG, greater accumulation of grams per day, Maravilla= 34 g day and HB-8= 42 g day 2B-5= 37g day), it could be considered that this material is precocious, since precocity is a relative characteristic and it is necessary to consider it within the production cycle, in places where the normal cultivation cycle goes from 140 to 180 days for the crops that are established in the area, this material shows to be of medium size with an average of two meters in length and a cycle of 140 days, its average yield is 3 205 kg ha^-1.

It is also clearly observed the division into two groups according to their precocity at the days of maturity where the following is shown: early cycle (138-143) medium cycle (151-166). For plant height, they are classified as low for the witnesses, Olga, Eva, K-93 and Zoya (100-162 cm), medium height with a height ranging from 230 to 236 for hybrids HB-8, Chinatan and 2B- 5. The tall one was Maravilla with 309 cm height, whose characteristic complicate the tasks of data collection and cutting (Table 3).

The variables involved with the yield, such as the fruit weight of the first bunch or generation one, the length of the first cluster and the weight of one thousand seeds showed values ​​that reflect the potential of the materials 2B-5 and HB-8 which responded with favorable yields in both evaluated cycles, densities and schedules of irrigation under study were relevant to identify their adaptability when subjected to water stress and increased plant competition.

In some studies, the number of seeds per bunch is broken down into the number of fruits per bunch and the number of seeds per fruit; however, these two components are preferably analyzed together because there is an insignificant variation in the number of seeds per fruit (^{Machado et al., 2009}). The individual weight of the seeds plays an important role, they reported yields of 3 569 kg ha^-1 for the Durango region.

On the other hand, ^{Machado et al. (2009)} reported 34 to 143 fruits per bunch. In Morelos they found ranges of days to flowering from 40 to 72 dds. The results found in this investigation were, ranges of DF fluctuated from 99 to 149 dds, the differences reside in the altitudinal level and climate, being Morelos of warmer and dry climate (Table 3).

Variation between hybrids by dates in combined analysis

In relation to the number of fruits (CFR) and grains in each bunch (CGR), a similar trend was observed in the first generation of bunches; therefore, this correlation is a consequence of the fact that each fruit contains three carpels that can originate the same number of grains; however, sanitary, genetic and environmental factors can regulate the final amount of grain per bunch. (^{Falasca et al., 2012}).

According to the work done by ^{Amorim et al. (2001)}; ^{Moshkin (1986 a)} and ^{Weiss (1983)}, extreme conditions, cold (<14 °C) and heat (>41 °C), limit the formation and maturation of seeds, due to flower abortion and sexual reversion, increase the ratio of male flowers to female ones. However, the results obtained indicate that, under the conditions where the research was conducted and for the varieties evaluated, the maximum and minimum temperatures of the environment did not affect the formation or filling of the grain, as discussed above. The cycle that most favors the development of the crop for HB-8 and 2B-5 is A-W with a density of 23 000 plants ha^-1 and three irrigation risks (Table 4).

Variation between risks by dates

The evaluated materials behaved in a favorable way to the irrigation schedules applied, since they were maintained with yields above 3 000 kg ha^-1, the commercial controls for their part, the highest in yield was OLGA with 2 681 kg ha^-1 and the lowest performing ZOYA with 359 kg ha^-1.

Density

Variability was found in hybrids of castor oil plant, in terms of the morphological characteristics of the plant, as well as in plant height, beginning of flowering, physiological maturity and above all in yield. This allowed to differentiate groups, which presented different production potentials overcoming the witnesses involved; According to the height of the plant, Silva (2005) grouped the castor oil plant in three categories: low plants (<1.5 m), medium (1.5-2.5 m) and tall (> 2.5 m) therefore, in accordance with the results obtained in this investigation showed that the height in the experimental lot with low density (23 000 plants ha^-1) the range of height was much wider where it is oscillating between 113 to 268 cm in length; however, when increasing the density to (26 000 plants ha^-1) it was reduced to a lower range ranging from 201 to 302 cm in height (Figure 1), it did not show significant effects in yield since higher density favors competition by light, nutrients; that is, it reduces the accumulation of biomass in the vegetative organs, decreasing the possibility of nutrients to supply the demand generated by spikes, capsules and growing grains (^{De Souza et al., 2011}).

The height of the vegetable is important, since according to ^{Weiss (1983)}, manual or mechanical harvesting is difficult, as the clusters of higher orders are emitted at higher heights.

Similar results were found by ^{Beltrão et al. (2001)}, who found, when comparing two locations with different altitudinal levels (27 and 500 masl), that as the height above sea level increased, the evaluated varieties presented greater size or size of the plant. As indicated by ^{Tavora (1982)}; ^{Weiss (1983)}; ^{Moshkin (1986 a)}; ^{Amorim et al. (2001)}; ^{Falasca et al. (2012)}, on the effect of the temperature and the limit altitude for the development of the plant and seed production, the results obtained in the present work, showed that the castor oil plant was adapted to thermal ranges below and above those that, traditionally, have been considered propitious for their cultivation; likewise, in relation to altitudinal elevation, it is evident that castor oil plant can be planted up to 2 120 meters above sea level, with the use of materials adapted to that environment, without affecting its performance.

Analysis by agricultural cycle

First planting date autumn-winter 2015-2016

Table 5 and 6 shows high significance among hybrids for all the characters evaluated; however, for the interaction hybrids by density only difference in yield was observed which indicates that some hybrids tend to be affected by the high density.

In the comparison of means for this cycle it is observed that the hybrids HB-8 and 2B-5 resulted with higher yield in this upper cycle 3 000 kg ha^-1, the most outstanding witness is Olga with yield of 2 681 kg ha^-1. Regarding the evaluation of plant height, it is observed that there are variations between hybrids, the highest observed height is in the Maravilla hybrid with 347 cm, while the average height is HB-8, Chinatan and 2B-5 which oscillates between 229 to 239 cm, however, the four controls were kept low in height with heights of 107, 160, 177 and 184 cm (Table 7, 8 y 9).

In Table 10 it is observed that the Maravilla, chinatan and 2B-5 hybrids were more yielding than Maravilla and HB-8 for this spring summer cycle; however, they did not surpass the K-93 control who showed a yield of 3 924 kg ha^-1. However, it is necessary to specify that the excessive growth of the hybrid Maravilla, caused difficulties for the manual harvest, since the height of the seed collection, aspect that was indicated by ^{Rodríguez and Duque (2010)}.

This is considered a common condition for most of the plants of castor oil to medium height to high and although the commercial varieties reach between 0.9 and 3 m (^{Flemming and Jongh, 2011}), many producers prefer dwarf varieties that grow at an average height 1.6 m, with periods of sowing to harvest from 120 to 130 days, as well as the medium varieties that grow from 2 to 2.5 m, with periods of sowing to harvest of 150 to 240 days (^{Rodríguez and Duque, 2010}), aspect that should be paid attention.