CHAPTER 12
Rice Population Improvement in Bolivia

Roger Taboada Paniagua^[23]
René Guzmán Arnez^[23]
Juana Viruez Justiniano^[23]
Tadao Kon^[24]

Roger Taboada Paniagua

Abstract

To develop rice varieties adapted to Bolivia’s upland ecosystems, the national Centro de Investigación Agrícola Tropical, Santa Cruz, began its project for population improvement through recurrent selection in 1997. It first introduced genetic resources from other countries, specifically three populations: CNA-7, PCT-4 and PCT-5, followed by populations PCT-7, GPCT-9 and PCT-11. Work was separated into two phases: (1) population characterization from which CNA-7, PCT-4, PCT-7 and PCT-11 were selected; and (2) population improvement and line extraction for developing new varieties. Population PCT-7 was chosen as the base to create a new population for favourable upland and irrigated conditions. Parallel with population characterization, fixed and segregating lines were introduced from the collaborative project on population improvement carried out by CIRAD and CIAT in Colombia. This chapter describes how the populations were managed, how fixed lines were obtained and how these genetic resources were evaluated. Preliminary results are positive; and lines developed from introduced populations are already permitting the planning of new varietal releases in 2004.

Resumen

En 1997 el Centro de Investigación Agrícola Tropical (CIAT-Bolivia) en Santa Cruz, Bolivia, inició el proyecto de mejoramiento poblacional del arroz a través de selección recurrente El objetivo era encontrar variedades adaptadas a las condiciones de secano para la producción de arroz en el país. El proyecto comenzó con la introducción de recursos genéticos existentes en otros países, es decir, se introdujeron tres poblaciones (CNA-7, PCT-4 y PCT-5), y después se adicionaron PCT-7, GPCT-9 y PCT-11. El trabajo se dividió en dos etapas: la primera, la caracterización de las poblaciones donde se seleccionaron las CNA-7, PCT-4, PCT-7 y PCT-11 y la segunda, el mejoramiento de las poblaciones y la extracción de líneas para el desarrollo de nuevas variedades. La población PCT-7 se seleccionó como base para formar una nueva población para condiciones de secano favorecido y riego. Paralela a la caracterización de las poblaciones, se introdujeron líneas fijas y segregantes del proyecto de mejoramiento poblacional colaborativo CIRAD/CIAT, Colombia. En este capítulo se muestran cómo se han manejado las poblaciones, cómo se obtuvieron líneas y cómo se realizaron las evaluaciones de esos recursos genéticos. Los resultados preliminares son positivos y el desarrollo de líneas a partir de las poblaciones introducidas ya permite planear el lanzamiento de nuevas variedades en el 2004.

Introduction

The use of conventional breeding methods has given Bolivian farmers new varieties, some of which - for upland conditions - are described by Guzmán (1995). Taboada et al. (2000) discuss achievements of the Bolivian rice programme, giving recommendations for the irrigated system. Increasing yields and improving the quality of new varieties have become the focus of the rice genetic improvement programme of the Centro de Investigación Agrícola Tropical, Santa Cruz, Bolivia (CIAT-Bolivia). To reach its goals, CIAT-Bolivia has had to confront difficulties, especially that of increasing productivity, perhaps because of the narrow genetic base of the resources available to the Centre’s programme (Taboada et al., 2000). This finding corroborates those of Cuevas-Peréz et al. (1992), who found that the commercial varieties planted in Latin America had a narrow genetic base.

Information received from farmers likewise indicates a constant demand for more productive materials. To respond to this demand, CIAT-Bolivia is seeking alternatives to make its genetic improvement programme for rice more dynamic and better able to exploit the genetic resources available in other improvement programmes found in Latin America. In 1997, Bolivian researchers participated in a consultative training event offered by Embrapa Arroz e Feijão, which emphasized the possibility of employing a new improvement methodology for the rice crop, and discussed the use of population improvement through recurrent selection.

The training event presented data on the advantages and disadvantages of applying the methodology to rice. With regard to the former, discussion included the possibility of creating and managing populations of wide genetic bases; the development of lines that theoretically could surpass the yield potential of commercial varieties currently used in Latin America, and the independence that this strategy offers by reducing the constant need to introduce germplasm from other countries and conduct manual crossing.

These positive aspects encouraged CIAT-Bolivia researchers to introduce germplasm and begin using population improvement in the country. This chapter aims to present the advances made so far by the rice genetic improvement programme with this methodology.

Characterizing introduced populations

Because the availability of rice genetic resources is so limited in Bolivia, in 1997, CIAT-Bolivia began its programme of rice population improvement through recurrent selection by introducing populations of wide genetic bases from the Rice Project executed by CIAT-Colombia in collaboration with CIRAD (France) and Embrapa Arroz e Feijão (Brazil).

The populations CNA-7, PCT-4 and PCT-5 were introduced and The data obtained led to the first two populations being chosen to continue with the process of population evaluation and improvement. The third population was discarded. Later, in 1998, population PCT-7 and the gene pool GPCT-9 were introduced for favoured upland conditions and irrigation, respectively. The results obtained from the first population were more favourable for our conditions; hence, it was selected as the source of male sterility to create a new local population (Taboada et al., 2000).

Table 1. Results of the evaluations carried out for seven agronomic traits in rice populations PCT-4, CNA-7 and PCT-11, 1999/00 season, Agricultural Experiment Station-Saavedra (EEAS), CIAT-Bolivia.

a. Number, SD and CV refer to number of plants evaluated, standard deviation and coefficient of variation, respectively.

b. White belly scores based on IRRI’s standard evaluation scale (1988). characterized on a preliminary basis.

In the 1999/00 season, when population PCT-11, which was developed in Colombia, was already introduced, the populations PCT-4, CNA-7 and PCT-11 were all characterized in detail. Available results (Table 1) indicated that the three populations adapted well to local conditions when planted in the upland ecosystem. With these data, the programme’s researchers used these populations as primary sources of germplasm for their strategy of population improvement. Because these sources were segregating for the male-sterility gene, genetic improvement was made easier.

To evaluate the three populations, 2000 plants per population were planted in the field over two seasons at 1000 plants per season. Their characterization - based on the sample sizes established by Badan et al. (1998) - was carried out on a sample of 200 plants each. The trials were conducted at the Agricultural Experiment Station at Saavedra (EEAS, its Spanish acronym). The following agronomic traits were evaluated: number of days to 50% flowering, plant height, number of panicles per plant, number of grains per panicle, grain length, white belly and reaction to diseases. Except for the last trait, the results of the characterization are summarized in Table 1.

Population CNA-7, on average, took 106 days to 50% flowering. Even so, it was relatively easy to obtain early maturing plants in this population, which also presented plants flowering at 81 days, a standard deviation of 6.6 days (Table 1). Plant height averaged at 129 cm, varying between 84 and 187 cm. Data on lodging (not presented) in this population, which was characteristically tall, indicated that 32% of the plants were susceptible.

Populations PCT-4 and PCT-11 had shorter cycles and were not so tall, averaging 100 and 98 days, and 105 and 98 cm, respectively. (These populations were intermediately tall and are destined for the mechanized farming system).

Because the numbers of panicles per plant and of grains per panicle are two important components of yield, we decided to measure these traits to have an idea of the potential of these populations. Population PCT-4 had the largest average number of panicles per plant, but the smallest average number of grains per panicle. In contrast, population PCT-11’s performance was reversed, that is, it had the smallest average number of panicles per plant and the largest average number of grains per panicle (Table 1). This relationship, common in rice, makes developing lines with increased yield potential difficult. However, as the ranges of variation for the three populations are wide for both traits and the standard deviations are high, recurrent selection is expected to enable the accumulation of genes that contribute to the plant breeders’ direction of interest, that is, increasing the average number of panicles per plant and of grains per panicle.

To respond to the demands of the national market, which requires long translucent grains, we determined the distribution of grain length and white belly within the populations. Table 1 shows that the three populations possessed means and ranges of variation that were desirable for grain length. However, the standard deviations were very low, which will surely make advances in selection difficult. With respect to white belly, the means were also within the desirable range, with values of less than 3 according to the IRRI scale of evaluation (1988), and presenting good variability for selection.

Figure 1. Strategy of germplasm flow adopted by the CIAT-Bolivia rice genetic improvement programme. Both the integration of population improvement strategies and line development are considered for varietal production.

Table 2. Flow and number of plants evaluated and selected in rice populations PCT-4, CNA-7 and PCT-11.

a. In the 1998/99 cropping season, population PCT-7 and gene pool GPCT-9 were also characterized.

b. X indicates population was characterized.

Although this study does not present data, general observations indicated that all the populations performed well with respect to the major diseases of the rice crop in Bolivia (leaf and neck blasts, helminthosporiosis, leaf scald and grain discoloration). Population PCT-11 had the highest number of susceptible plants to the two last diseases.

Introduced populations and their management

The characterization of the introduced populations carried out in 1997/98 enabled us to determine, in a preliminary manner, their phenotypic characteristics and performance under upland conditions in Bolivia. This led to initiating the process of population improvement when the first recurrent cycle was carried out on the basis of mass selection for both sexes, using the criteria previously described and evaluated. Later, the three populations, selected and characterized in 1999/00, fulfilled the proposed stages (Figure 1): fertile plants selected; S_0:1 generation advanced to S_0:2; S_0:2 families evaluated and selected; and the best S_0:2 families recombined with remnant S_0:1 seed.

During the 1999/00 summer, in the first stage of population improvement based on S_0:2 families, 96 fertile S₀ plants from population PCT-4 were selected, 86 from CNA-7 and 88 from PCT-11. These materials were then advanced from the S_0:1 generation to S_0:2 in the 2000 winter, under irrigated conditions at the EEAS. The families were then evaluated during the 2000/01 summer at the EEAS, with 38, 31 and 32 S_0:2 families being selected from populations PCT-4, CNA-7 and PCT-11, respectively. Later, with the selected material, a balanced mixture was formed with the remnant S_0:1 seed for recombination (Table 2).

Because, in Bolivia, only one generation per year can be developed, during the second semester of 2001, collaboration was requested from CIAT-Colombia to speed up the process. Three bags of seed were sent, each representing one population, for recombination. This stage was carried out by planting each population in isolation and on two different planting dates to facilitate and guarantee as much crossing as possible. Later, all male-sterile plants were harvested from each population without selection to form a balanced mixture under irrigated conditions. All male-sterile plants were individually harvested (about 700 plants from each population) to generate new populations with two cycles of recurrent selection and more adapted to Bolivian conditions, after the first mass selection.

This permitted completing the second recurrent cycle in populations PCT-4 and CNA-7, giving rise to new improved populations for the agronomic traits considered, and named according to the nomenclature recommended by Chatêl and Guimarães (1995), as follows (‘CA’ refers to the Spanish características agronómicas, that is, to agronomic traits’):

• PCT-4\CA\1\1,CA\1
• CNA-7\CA\1\1,CA\1

In contrast, population PCT-11 completed one recurrent cycle to obtain the improved population PCT-11\CA\1\1 Table 2).

This information helped focus the improvement of the populations according to traits such as grain quality (white belly), resistance to fungal diseases and yield. To do so, populations PCT-4 and PCT-11 were established to work with farmers possessing technologies for the mechanized farming system. With respect to CNA-7, the possibility of covering the manual system of small farmers was considered. This germplasm is one of the few alternatives for this production system.

We point out that in all stages where the segregating populations were planted in the field, plants and/or lines were chosen to produce, through the pedigree method, advanced materials for the development of commercial varieties.

CIAT-Bolivia’s rice genetic improvement programme considers that the population improvement method through recurrent selection is a good alternative for responding to the demands received by the Centre. This is based on the methodology permitting results to be obtained in the medium and long term such as populations with a high allelic frequency for traits of interest and segregating lines that can result in new varieties for the upland ecosystem. Likewise, population improvement can be combined with line development, as described in Figure 1.

Table 3. Flow of rice germplasm introduced from the CIRAD/CIAT Rice Project, Colombia (1998-2001), and the process of selection for the upland conditions of Bolivia (2001/02).

Because of the climatic conditions in Santa Cruz, only one generation per year can be carried out, and each cycle of population improvement, using the proposed improvement methodology, requires 4 years (Figure 1). However, support from CIAT-Colombia has helped reduce this period to 2 years, as described in Table 2.

Strategies for developing lines by exploiting populations

The elements related to population improvement have been discussed in the previous sections. Now we will comment on the process of obtaining improved lines from segregating generations available during population improvement. The process begins with individually selecting 13 fertile S₀ plants and advancing them to fixation through genealogical or pedigree selection. After trials for evaluating and selecting S₂ families, carried out during the 2000/01 season, 236 S₃ lines (individual plants) were selected, of which 89 came from PCT-4, 67 from CNA-7 and 80 from PCT-11. This germplasm was developed to S₅ through the conventional method of line development to later continue with evaluation and selection in yield trials.

Yield trials and evaluation of S_0:2 families from population improvement were carried out at the sites of Saavedra, San Pedro, Yapacaní and San Juan, which represented the rice-producing regions of the Department of Santa Cruz. From 1998 to 2002, 856 fixed and segregating lines (Table 3) were introduced from the CIRAD/CIAT Rice Project and, from the 2001/02 harvest, trials were conducted in all phases of improvement.

Table 4. Means of evaluations for number of days to 50% flowering, plant height and white belly of the S_0:2 families of rice populations PCT-4, CNA-7 and PCT-11, and local checks, 2000/01 season.

a. F refers to S_0:2 families evaluated; FS refers to S_0:2 families selected.
b. Average scores according to IRRI’s White belly scale (1988).

In the introduction trial of the 2001/02 harvest, 173 lines were evaluated; in the preliminary comparative yield trial, 14 genotypes; in the advanced 9; and in the regional adaptation, 6 lines (Table 3). All these materials originated from the populations being improved under the recurrent selection methodology, and thus received the code ‘SR’ (Spanish abbreviation for ‘recurrent selection’). To select them, the following aspects were considered: tolerance of diseases (leaf and neck blasts, helminthosporiosis, leaf scald and grain discoloration), short or intermediate cycle (100-135 days), good grain quality (little incidence of white belly and high milling yield) and a yield potential that is similar to or better than that of the local checks.

Results of strategies for selecting lines

As already mentioned, during the 2000/01 harvest, an evaluation was conducted for 96 S_0:2 families selected from population PCT-4, 86 from CNA-7 and 88 from PCT-11. The statistical design of Federer’s augmented blocks was used following Zimmermann (1997). The families were selected, based principally on the following agronomic traits: grain health and quality; number of days to 50% flowering and white belly, with the last having the most influence over selection. Results are shown in Table 4. The mean for white belly for all S_0:2 families evaluated in population PCT-11 was 2.8, whereas the selected families averaged 2.0, that is, the value dropped by 0.8 (28.6%). For population PCT-4, incidence of white belly dropped by 0.6 (19.3%) and for CNA-7 by 0.7 (23.3%)

Table 5. Means of yield (kg ha^-1) of promising rice lines developed through recurrent selection, CIAT-Bolivia, 2001/02 season.

a. Mean across four sites: Saavedra, Yapacaní, San Pedro and San Juan.

On analysing all the yield trials conducted by the programme, promising results can be seen for the genotypes introduced in 1999 from the CIRAD/CIAT Rice Project. These materials were evaluated in introduction trials in observation plots in 1999/00, with preliminary comparisons made in 2000/01 and advanced comparisons in 2001/02. The last two phases were evaluation trials at different sites. In the next harvest (2002/03), the last phase of the yield trials were evaluated. Table 5 presents the results of the 2 years, in which genotypes SR99343, SR99380, SR99345 and SR 99344 stood out for their high yield potential, besides being significantly superior to the local varieties used as checks (Jasayé and Tutuma). These lines possessed good grain quality and were tolerant of the major fungal diseases found in the rice-growing areas of Santa Cruz. These promising materials came from population PCT-4, thus indicating the potential of both the population and the strategy for developing new rice varieties for upland planting.

Prospects for the methodology and future plans

The positive preliminary results obtained from using rice population improvement in the country indicate the achievement of some of the objectives established at the beginning of this work. The programme now possesses broad sources of genetic resources adapted to the predominant ecosystems of Bolivia.

When compared with local materials in different yield trials, the lines developed by CIAT-Bolivia from the introduced populations and improved for upland conditions and subsistence rice farming demonstrated positive performance. In the near future, the programme will be able to provide farmers with new varieties. The good performance of these lines with respect to disease resistance and productivity should also be highlighted.

However, implementing the strategy as proposed in Figure 1 requires an administrative vision of supporting the work, not only in the short term, but also over the medium and long term. At first, population improvement is centred on accumulating favourable genes and on increasing possibilities of obtaining better genotypes at each recurrent cycle.

The CIAT-Bolivia rice programme will continue managing the three populations used so far, and is planning to conduct the following activities in the next few years:

• To introduce a population for irrigated conditions

• Starting in 2003, to form a population for favoured upland conditions, based on population PCT-7, by introducing parental materials that are adapted to our conditions

• In 2004, to release two upland varieties from populations managed through recurrent selection.

References

Badan, A.C.C.; Geraldi, I.O.; Guimarães, E.P. & Ospina, Y. 1998. Estimativa do tamanho efetivo de amostra ideal para caracterizar uma população de arroz para as características peso de panículas e peso de 100 grãos. Genet. Mol. Biol., 21: 246.

Châtel, M. & Guimarães, E.P. 1995. Selección recurrente con androesterilidad en arroz. Cali, Colombia, CIRAD-CA & CIAT. 70 pp.

Cuevas-Pérez, F.E.; Guimarães, E.P.; Berrío, L.E. & González, D.I. 1992. Genetic base of irrigated rice in Latin America and Caribbean, 1971 to 1989. Crop Sci., 32(4): 1054-1059.

Guzmán, R. 1995. Nuevas variedades de arroz para secano en Santa Cruz, Bolivia. In Informe INGER-América Latina 1994; Variedades de arroz liberadas en América Latina y el Caribe en 1993 y su variabilidad genética, pp. 15-19. Cali, Colombia, International Network for Genetic Evaluation of Rice for Latin America (INGER-Latin America).

IRRI (International Rice Research Institute). 1988. Standard evaluation system for rice. 3rd ed. Los Baños, Philippines. 54 pp.

Taboada, R.; Guzmán, R. & Hurtado, J. 2000. Mejoramiento de arroz en Bolivia: Caracterización de poblaciones y uso del mejoramiento poblacional. In E.P. Guimarães, ed. Avances en el mejoramiento poblacional en arroz, pp. 287-296. Cali, Colombia, CIAT.

Zimmermann, F.J.P. 1997. Estadística aplicada a la selección recurrente. In E.P. Guimarães, ed. Selección recurrente en arroz, pp. 67-75. Cali, Colombia, CIAT.