Performance of eleven introduced improved lowland rice varieties in the northern Savanna zones of Ghana

From 2007 to 2010, we evaluated eleven introduced improved rice materials, mainly lowland NERICAs and Thailand varieties for yield and other farmers’ preferred agronomic traits compared to Gbewaa (aromatic local check) and Digang (non aromatic local check). The trials, each year, were laid out in a Randomized Complete Block Designed (RCBD) with four replicates at two locations (Nyankpala and Salaga) in Northern Region, Ghana. Six cultivars namely, WAS 163-B-5-3, WAS 122-13-WAS-10-WAR, L2-4, PERFUME IRRIGATED, LONG GRAIN ORDINARY 2 and EXBAICA yielded significantly higher (p<0.5) with mean yield advantage range of 21 to 73% over Gbewaa, 16 to 66% over Digang, adapted to the target environment and tolerant to the major biotic stresses. These materials include two aromatic varieties which could serve as substitutes for Gbewaa (the market aromatic quality) and give farmers a choice to meet the existing market demand for aromatic rice. 
 
   
 
 Key words: Rice evaluation, Savanna agro ecologies of Ghana, NERICAs, improved varieties.


INTRODUCTION
Rice has successfully become a staple food in Ghana.This is a result of the rapid change in lifestyle and food habit; particularly in urban centers (Nyanteng, 1987;Balasubramanian et al., 2007).The per capita rice consumption has risen from average of 17.5 kg in 2001 to 22.6 kg in 2004 and is estimated to reach 63 kg in 2015 (JICA, 2007).
Ghana continues to import about 60% of its rice needs and consequently spends about US$ 450 million annually on imports (MOFA, 2010;Mohapatra, 2012).Efforts to increase domestic rice production will contribute directly to foreign currency saving and balance of payment, in addition to improving farmers' incomes.Increased productivity from increased yield per unit area through the use of high yielding locally adapted varieties and good agronomic practices is the most viable strategy to reduce dependency on importation (MOFA, 2009).
Ghana mostly depends on the Consultative Group on International Agricultural Research (CGIAR) centers, that is, Africa Rice Centre (AfricaRice) and International Rice Research Institute (IRRI) for its improved breeding materials (MOFA, 2009).
The Participatory Varietal Selection approach (PVS), which aids easy access of promising materials to farmers (AfricaRice, 2010), is mostly employed in evaluating the materials.Promising ones tested at multi-locations and on-farm are recommend for release as varieties.The savanna zones of Ghana which constitute the three Northern regions (that is, Northern, Upper East, and Upper West) produce over 60% of the local rice (JICA, 2007).The Savanna Agricultural Research Institute (SARI), one of the 13 institutes of the Council for Scientific and Industrial Research (CSIR) is mandated to conduct research and introduce improved technologies to enhance agricultural productivity in this zone.
As part of its mandate, new higher yielding rice varieties are to be introduced to farmers to enable them improve their productivity.This paper presents the performance of introduced lowland NERICAs and some Thailand materials in the Savanna agro ecological zones of Ghana.

MATERIALS AND METHODS
From 2007 to 2010, we evaluated at Nyankpala and Salaga five lowland NERICA Varieties, 5 introductions from Thailand, one introduction from University of Ghana Research station (Kpong) and two local control varieties.The five Lowland NERICAs were the most preferred farmer selections out of 30 entries during a previous Participatory Variety Selection (PVS-R) in 2004 (SARI, 2010).The introductions from Thailand and Kpong were included in 2008.The list of varieties used for the multi-location tests are presented in Table 1.

Experimental sites
The materials were tested at CSIR-SARI main station at Nyankpala and sub-station at Salaga.The Nyankpala site which falls within the Guinea savanna agro ecological zone is located 16 km West of Tamale the capital city of northern region of Ghana.The area lies in latitude 9° 25' 41" N and longitude 0° 56' 42" W, altitude 183 m above mean sea level.Annual average rainfall is about 1000 mm and mean temperature of 39°C (Kasei, 1988).The Salaga site falls within the west savanna transitional agro-ecological zone and is located 5 km south of Salaga the capital city of East Gonja district.
The area lies in latitude 8° 33' 25" N and longitude 0° 31' 04'' W, 156 m above mean sea level.Annual average rainfall is 1100 mm and mean temperature 27.9°C (Kasei, 1988).Weather conditions during the experimental years at Nyankpala is presented in Table 2.

Characteristics of the experimental soil
The characteristics of the experimental soils at the two sites are presented on Table 3.Total N was determined by the Kjeldahl procedure (Tel and Hegatey, 1984), available P by the Bray 1 extraction solution procedure (Bray and Kurtz, 1945) and exchangeable K by 1.0 M neutral NH4OAc solution (Black, 1965).According to the physical and chemical characteristics, the soils can be described as deep sandy loam to clay loam (Buri et al., 2012).The pH was 4.8 at the 0 to 15 cm level and 4.9 at 15 to 30 cm.Compared to Buri et al. (2012), the levels of N, P and K in the soils were considered to be low.

Field preparation and experimental design
The experimental sites each year were ploughed around the first week of June and harrowed a week or two after (at the middle of June).A Randomized Complete Block Design (RCBD) with four replicates and a unit plot size of 3 m x 5 m was used.

Planting and cultural practices
Planting was done by dibbling seeds in a space of 20 cm × 20 cm between rows and hills.One seedling was allowed per hill three weeks after sowing.The recommended rate of compound fertilizer N-P-K of 60-60-30 kg/ha was applied in two split applications (Ragasa et al., 2013).Three weeks after each planting date, the basal fertilizer of 30-60-30 kg/ha NPK (grade 15-15-15) (300 g) and triple superphosphate fertilizer (98.7 g) were applied per experimental plot of 15 m 2 .The remaining nitrogen was applied as top dressing at seven weeks after planting using sulphate of ammonia (214 g/plot).Weeds were initially controlled with preemergence herbicide (Pendimethaline 400 g/l; Alligator) at 3.2 L/ha a day after planting and then by hand weeding.Rice was harvested and threshed manually at maturity using sickles and beating on tarpaulin respectively.After measurement of grain moisture, weight

Traits evaluation
Traits evaluated for the study include grain yield, plant height, number of tillers, number of panicles, days to 50% flowering, maturity days, 1000 grain weight, disease scores for blast, brown spot and leaf scald based on IRRI's Standard Evaluation System for rice (IRRI, 1996).

Statistical analysis
Analysis of variance was first performed (SAS Statistical Package) followed by LSD for mean comparison and separation when necessary.

RESULTS
Results of selected parameters measured in 2007 are summarized in Table 4.The recorded yields for the two locations; Nyankpala and Salaga in 2007 are also presented in Table 5.The yield performance for the years 2008, 2009 and 2010 are presented in Table 7.

DISCUSSION
Generally, the recorded yields were low compared to previous report (JICA, 2007).The yield ranged between 0.7 and 5.5 t/ha for Nyanpkala and 0.8 to 4.5 t/ha at Salaga over the years.Gbewaa, the aromatic check recorded average yield of 2.28 t/ha (Table 7) over the years which is just about 40% of its potential yield of 6 to 7 t/ha (SARI, 2008).Intermittent moisture stress could account for the recorded low yields since experimental fields were rain fed and the plants occasionally suffered from severe moisture stress during the various seasons especially at Salaga in 2010 where the recorded yields were particularly low (Table 7).Terminal season drought is the major abiotic constraint of rice production in the savanna regions (MOFA, 2009).Digang, the non aromatic check averagely performed better (2.4 t/ha) than Gbewaa (2.28 t/ha) (Table 7).
Digang is known to tolerate drought better than Gbewaa.Gbewaa is reported to have problem  (Acheampong, 2010).Planting early and medium duration varieties has been the reliable strategy to escape the effect of terminal season droughts.The selected materials are of medium duration (ranging from 108 to 130 days) (Table 6) which makes them suitable for cultivation in the region.However, to ensure stable yields and sustainable cultivation of these materials in the drought-prone savanna zones, effort should be made to improve the level of drought tolerance of these materials.Analysis of variance across the two locations showed no significant genotype by location interaction indicating that the selected materials can be grown across the target environment.
The selected materials have plant height range of 96 to 126 cm, (Table 6), mostly semi dwarfs to avoid lodging.The materials also have adequate tolerance to brown spot, leaf blast and leaf scald (Table 6); the most prevalent diseases in the tested environment.Screening against these diseases was incorporated in the evaluation process right from the observational nursery scoring each year based on the Standard Evaluation System (SES) of rice by IRRI (1996).The materials also have appreciable levels of tolerance to African gall midge; the most prevalent insect pest (Table 6).
The five introductions from Thailand were included in 2008 because of the market demand for aromatic rice.The Ghanaian market has high preference for intermediate amylose, long slender aromatic rice grains, with these characters accounting for over 40% increase in price quotations (Asante, 2005;Minkah, 2007;Anang et al., 2011).Gbewaa rice is cultivated by majority of Ghanaian farmers because of its aromatic quality.However, it is susceptible to blast and Rice Yellow Mottled Virus (RYMV) (Asante, 2012).Digang (the non aromatic check) is resistant to these two diseases but not preferred because of its poor grain quality (Asante, 2012).

Conclusion
Six of the test materials (WAS 163-B-5-3, WAS 122-13-WAS-10-WAR, L2-4, PERFUME IRRIGATED, LONG GRAIN ORDINARY 2 and EXBAICA) outperformed the local popular checks (Digang and Gbewaa) and are currently undergoing on-farm testing pending possible recommendation for release as varieties.Two aromatic varieties (PERFUME IRRIGATED and LONG GRAIN ORDINARY 2) could serve as substitute for Gbewaa and give farmers a choice to meet existing demand for aromatic rice.

Table 1 .
Varieties and sources used for the Multi-location trials.
*Released varieties used as local check; ** Introduced in 2002 and released as a variety in 2009.

Table 2 .
Summary of weather data for May to October at Nyankpala for the experimental years.

Table 3 .
Physical and chemical characteristics of soils used.

Table 5 .
Yield (t/ha) of promising varieties at the two sites in 2007.

Table 6 .
Agronomic characteristic of Lowland NERICAs and introduced varieties from Thailand in 2008.

Table 7 .
Yield (t/ha) of varieties at the two sites in2008, 2009, and 2010.Two of the Thailand introductions (PERFUME IRRIGATED and LONG GRAIN ORDINARY 2) proved suitable for cultivation in the savanna ecology.Other interesting characteristics of the selected materials include high tillering ability, large panicle size and long grains preferred by the Ghanaian market.