In most developing countries where barley is cultivated, the crop predominates in areas where other rainfed cereals are unprofitable or not possible to grow. This is due to climatic stresses, among which drought is by far the most important. Therefore, breeding barley cultivars more resistant to drought than those currently available is one of the most important objectives of the barley-breeding program at ICARDA.
Water is becoming an increasingly scarce resource—about one-third of the human population does not have sufficient water; and given that about 70% of available water is used in agriculture, breeding for drought resistance has to be seen as one of the best ways, and possibly one of the cheapest ways, of saving water globally.
Studies show that water is used inefficiently in both irrigated and rainfed agriculture. There are two main ways of increasing water-use efficiency in agriculture. One is to
In discussing with researchers the rainfall distribution and risk in her barley production, farmer Juri Aboud used stones of different sizes as visual aids. The survey results indicated that farmers preferred a stable variety over a high-yielding but unstable one.
Barley is a good model plant to study physiological, genetic, and breeding aspects of drought resistance. In many countries the crop is of great economic importance in areas where drought and other abiotic and biotic stresses make the cultivation of other rainfed crops impossible or uneconomical.
Two contrasting approaches have been followed in breeding crops for drought tolerance. The first uses selection under optimum growing conditions, and is based on the assumption that increased yield potential will have a carry-over, or spillover, effect when the improved cultivar is grown under less favorable conditions. The second uses direct selection in the presence of drought, i.e., in the target environment, and can take two forms: selection for physiological or developmental traits (also called analytical breeding), and direct selection for grain yield (also called empirical or pragmatic breeding). Environment here is defined as the complex of climate, soil type, soil depth, soil fertility, agronomic management, etc.
The first approach has failed to produce convincing results. In fact, drought continues to affect agricultural production negatively worldwide despite spectacular increases in crop yield potential obtained through breeding under optimum conditions. An example of the absence of spillover effect is offered by wheat production in Syria, which in the period 1984-2000 (the first improved cultivar was released in 1983) increased by a staggering 50 kg/ha per year, to the point where Syria is now a wheat exporter. However, this large increase in yield potential of new cultivars was achieved under irrigated conditions and had no significant effect on yield under rainfed conditions, even though the adoption level of improved cultivars has reached almost 90%.
Breeding for drought resistance based on putative traits (traits associated with drought resistance, but easier to select for than grain yield) has been and still is very popular. Traits that have been investigated include (1) physiological/biochemical traits, such as proline content, stomatal conductance, epidermal conductance, canopy temperature, relative water content, leaf turgor, abscisic acid content, transpiration efficiency, water use efficiency, carbon isotope discrimination, and retranslocation; and (2) developmental/ morphological traits, such as leaf emergence, leaf area index, leaf waxiness, stomatal density, tiller development, flowering time, maturity rate, cell membrane stability, cell wall rheology, and root characteristics. In the case of barley, we have found that the traits more consistently associated with higher grain yield under drought include growth habit, early growth vigor, earliness, plant height under drought, long peduncle, and short grain filling duration.
While the analytical approach has been very useful in developing an understanding of which traits are associated with drought tolerance and why, it has been less useful in actually developing new cultivars with improved drought resistance under field conditions. This is because, under field conditions, drought varies in timing, intensity and duration, and therefore it is the interaction among traits that determines the overall crop response to the variable nature of drought stress, rather than the expression of any specific trait. Recently, a similar level of complexity has been revealed in analysis of drought responses at the molecular level. In addition, drought is often associated with other abiotic stresses, particularly temperature, either high or low, as well as biotic stresses, and the way in which these various stresses combine and occur.
Although breeding for drought resistance based on direct selection for grain yield in the target environment (empirical or pragmatic breeding) appears intuitively the most obvious solution, this approach faces two major problems: the precision of the yield trials conducted in dryland conditions, and the existence of several target environments, each characterized by its own specific type of drought and combination of stress.
The first problem has been largely overcome by the considerable progress that has been made recently in both experimental layout and statistical analysis. It is now possible to conduct trials outside the closely controlled conditions of the research station without losing precision.
The problem of target environments is typical of areas where the risks of crop loss and of low yield are so high as to discourage the use of external inputs. This low use, or no use, of inputs has the effect of maximizing environmental differences between different areas and between different cropping seasons within the same area. Breeders working in unfavorable environments must account for another challenge—large genotype-by-environment interaction, which makes finding a broadly adapted variety more difficult. To account for this, the ICARDA barley breeding program has used selection for specific adaptation (adaptation to a specific harsh environment) as a strategy to breed for drought resistance. Critics point out that this strategy is usually associated with a reduction in yield potential under favorable conditions. The matter must be considered in its social context, and in relation to the difference between adaptation over space versus adaptation over time. For example, Australian farmers prefer maximizing yield in favorable years, while for North African and Near Eastern farmers, yield in very poor years is more important.
Selection for specific adaptation has entailed evaluating early segregating populations (populations still showing much variability between plants) in dry sites, such as Breda and Bouider in Syria. The most important results of this choice have been the re-evaluation of the role of landraces in breeding for drought tolerance, and the discovery of the importance of the wild progenitor of cultivated barley, Hordeum spontaneum, as a source of resistance to extreme levels of drought. This was evident during the drought of 1987 when two lines of Hordeum spontaneum were the only survivors in the breeding nurseries grown at a very dry site, which received just 176 mm of rainfall.
In the early 1990s the concept of breeding for specific adaptation was incorporated into the development of germplasm for international distribution, using a decentralized breeding approach. By this we mean the development of segregating populations from targeted crosses and their distribution to each of several target regions for selection and testing.
There are some major implications of adopting decentralized breeding as a breeding approach in international breeding programs. First, national programs generate more varieties, each adapted to specific conditions, including the different types of drought characterizing each country, and, second, the superior performance of the varieties developed for low-input and less-favored lands will not depend on agronomic practices that require large amount of inputs. A breeding program based on this approach is less likely to endanger biodiversity and the environment.
Although the concept of decentralized breeding is usually well received by national programs, decentralization per se does not necessarily respond to the needs of resource-poor farmers in less-favored areas. If decentralization means simply a shifting of breeding work to outlying stations, and if the outlying stations are not representative of a country’s harsh marginal areas, as is often the case, then nothing much is achieved. To exploit potential gains from specific adaptation to low-input conditions in general, and to drought in particular, breeding must be shifted from research stations to farmers’ fields.
Although
decentralization and farmer participation are unrelated concepts, decentralization
to farmers’ fields almost inevitably leads to the participation of
farmers in the selection process. In the case of the ICARDA barley program,
the idea of farmer participation in formal breeding programs arose originally
from a desire to conduct decentralized selection to exploit genotype ´
environment interactions and a desire to capitalize on farmers’ knowledge
about the crop, its specific uses, and its specific adaptation. In other
words, decentralized, participatory barley breeding represents an evolution
of a breeding strategy to improve drought resistance.
Progress has been slow, as expected, given the
large genotype ´ year interactions, and given the fact that drought
Participating farmers discuss with researchers the performance of their barley crop in a drought-hit field.
resistant lines can only display
their true value when a drought actually occurs. The yield advantage of
some of the highest yielding lines at a dry site in 1999 (200 mm rainfall)
amounted to 14% over the landrace and 38% over Rihane-03, which yielded
nearly 20% less than the local landrace in dry years and in dry sites.
Year
2000 saw the first real demonstration that the combination of 1) direct
selection in the target environment, 2) the use of landraces and H. spontaneum,
and 3) plant selections made by farmers, could increase the crop’s
resistance to extreme drought. In that year, total rainfall in most areas
of Syria was below average, and crop yields were severely affected. In
some areas, rainfall was so low that the crop failed to germinate, in
many others the crop failed to produce grain.
In trials in farmers’ fields in some very
dry locations in Syria (Tel Brack, Hassakeh province with 87 mm rainfall;
Jurn El-Aswad and Bylounan, Raqqa province, 121 and 87 mm, respectively;
and Bari Sharky, Hama Province, 130 mm), the only entries able to produce
grain and/or biomass were all derived from crosses with those two lines
of H. spontaneum identified in 1987 at Bouider. Grain yield was between
300 and 500 kg/ha and biomass yield was between 500 and 3000 kg/ha. These
yields appear very low, but to the farmers in these areas any yield is
very important because it reduces the economic impact of drought.
Data obtained during 2002 confirmed that decentralized, participatory plant breeding can be very powerful in adapting the crop to a wide range of specific environments, even within a limited geographical area.
Figure 1. Grain yield of barley lines selected by farmers in four villages in the dry areas of Syria compared with the local landrace Arabi Aswad (in green).