Coffee Leaf Rust (CLR) has been a disease of considerable economic significance in all coffee growing countries ever since it became an epidemic in the late 19th Century in Sri Lanka and India. CLR was reported first in 1861 from the wild Arabica coffee plants growing in the areas surrounding Lake Victoria in Kenya. On cultivated coffee the disease was first noted in the year 1868 in Ceylon. The crop losses caused by the disease are so high that within a decade Arabica coffee cultivation in that country was replaced by Robusta.,Similar.events took place in Indonesia also.

The disease was widespread in the coffee growing countries of Asia and Africa by 1920. In the Latin American countries growing coffee, it first appeared in the year 1970 and quickly spread to most of the South and Central American countries. At present, Hawaii is the only coffee producing country free of CLR. In India, the disease was first noticed on an epidemic scale in the year 1869. Enterprising 'coffee planters of India resorted to selection of tolerant types, introduction of hardy species and even hybridization of Arabica and the hardy species like Robusta and Liberica. The Folk Selection Era of coffee breeding has happened in our country and pioneer coffee growers like Stanley Jupp and LP. Kent and others are remembered even today for their contributions. The variety of developed materials like Hamiltons, Jacksons, Netrakonda and Chandrapore hybrids are worthy of mention in the history of coffee breeding.

Coffee Breeding

Arabica coffee (Coffea arabica L.) is a perennial plant ravaged by several diseases and pests. Of the various biotic adversaries of coffee, leaf rust disease caused by Hemileia vastatrix B. et Br. is the most important. Working with leaf rust and coffee in India in 1932, Mayne identified four physiological races of leaf rust on Arabica coffee. Important knowledge revealed by this study is the existence of physiological specialization in leaf rust 'and race-specific resistance in the coffee plant.

The pattern of inheritance of the resistance factors was not clear at that time. Using these meagre ingredients of knowledge researchers in India embarked up' on a systematic programme of evolving rust resistant coffee plants through breeding at the Mysore Coffee Experiment Station (1925; present Central Coffee Research Institute). In a review of the early breeding work in 1975, Srinivasan and Narasimhaswamy stated that, "coffee breeding is a very difficult problem". This is very true given the complexities added by the inherent nature of the coffee plant such as the tetraploidy and selfcompatibility of Coffea arabica versus the diploidy and selfincompatibility of most of the other species, chromosome homeologies across the wide array of diploid species and the tetraploid C. arabica, free inter-crossing of species within the diploid gene pool as well as between diploid and tetraploid species and the spontaneous occurrence of reasonably fertile diploid and tetraploid interspecific hybrids.

This scenario also presents wonderful opportunities for evolving coffee strains possessing special combinations of characters to cater to the preferences of consumers and connoisseurs through designed breeding programmes. Thus, despite difficulties, steady progress was achieved in coffee breeding all over the world leading to a global overproduction of coffee and consequent imbalance of demand supply situation. World production of coffee increased by about 14% in the past one and a half decade. Surplus in the world market precipitated in a price crisis of unprecedented dimensions. In this situation, producing special and specialty coffees under organic farming appear to be offering hope to the growers. Evolving coffee strains suitable to these grower and consumer preferences is a new challenge to the breeders. In the following sections, progress in coffee breeding research is reviewed to bring the stock of available knowledge into perspective and to plan for the future.

Breeding t~chniques and strategies for the commercially important C. arabica and diploid species like Robusta (c. canephora Pierre ex Froehner) are significantly di.fferent on account of the differences in their reproductive biology. While C. arabica is a tetraploid with selfcompatible reproductive system others are diploid with selfincompatibility and obligate crossfertilizing reproduction. This situation also creates barriers for transfer of genes of resistance available in the diploid species to Arabica. Present paper summarises the progress of research work on breeding of coffee for leaf lUst resistance in our country. The entire research can be divided into two major phases. The first phase was between 1925 and 1940 when primarily the indigenous materials collected from various coffee estates in India were exploited to evolve lUst resistant strains. The second phase started in 1953 with the introduction of exotic germplasm from various coffee growing countries and semiwild and wild genotypes from Ethiopia the homeland of Arabica coffee.

Early Indian Selections

In the course of assessing indigenous germplasm collections, one plant of the collection S.26 from Doobla Estate was found to manifest resistance to the races I and II of the rust fungus prevalent at that time. Self pollination derived progeny of this mother plant (S.288) was released to the industry for commercial exploitation in 1937, as Selection-I, considering the need of the hour. Later, research work at Centro Investigacao das Ferrugens do Cafeeiro (CIFC), Portugal revealed that this selection is a carrier of the gene SH3 that conditions its resistance. Even though S.288 was resistant to the prevalent races of the rust, it produces considerable quantities of triage (mis-shaped and malformed beans) that reduces the market value of the produce. To improve produce quality, S.288 was crossed with Kents variety that was known for quality. The progeny S.474 was found to manifest good vigour and resistance to the prevalent races. S.795 derived from this and released as Selection-3 (in 1946-47) was observed to be segregating in a 3: I Mendelian ratio of resistant and susceptible plants.

The SH3 gene that was responsible for the resistance of these materials was proposed to have been derived from the wild diploid species of West Africa, Coffea liberica Bull ex Hiern. However, the gene was, so far, not isolated from C. liberica. DNA marker studies revealed the presence of similar AFLP fragments in C~ liberica and the S.288 derivatives S.795 and S.1934. From the available information on markers and the stability of the transmission of this gene in breeding exercises, it can be inferred that SH3 originated by an initial homologous recombination between chromosomes of C. liberica and C. arabica followed by an intragenic recombination to create the unique gene that is found only in the early Indian coffee selections.

New Coffee Selections

The second phase of coffee breeding in India began in 1950s when a massive germplasm bank of over 400 cultivated and wild Arabica genotypes was established at CCRI. Ethiopian land races Agaro. Cioccie. Tafarikela. SI2Kaffa and Geisha were observed to be offering the potential of excellent quality as well as considerable resistance conditioned by the genes SH 1, SH2, SH4 and SH5 in different combinations. Produce quality of some of the lines such as Agaro and Cioccie was also found to be very good with a large frequency of A-grade beans. Ethiopian Arabicas Agaro, Cioccie and Tafarikela constitute the individual lines of Selection-4. Even though these materials offered an immediate sequel to the onslaught of race VIII (V2.35) that increased largely on the early selections (in 1960s), they are susceptible to race XXIV (VI.2.4) that has come to prevalence in the subsequent times. This exp~rience has provided the understanding that successive addition of single resistance genes or materials carrying a single new gene gives only short-lived respite and paved the way for experimenting with the high resistance found in the diploid species and tetraploid interspecific hybrids.

Interspecific Hybrids and Derivatives

In the year 1937, Robusta (S.274) and Arabica (Kents) were crossed in the intention of creating materials with Robusta type of disease resistance and Arabica type of quality produce. The resulting triploid was unstable and gave very small fruit set. This was backcrossed to the Kents parent to derive stable tetraploids. After two backcrosses, stable tetraploids with high resistance to the disease and good quality produce were generated and released as Selection-6 in early 1970s. Majority of plants (> 80%) in this selection are resistant to leaf rust in the field.

At about the same time, Hibrido de Timor (HDT) known for its resistance to all races of rust was introduced from CIFC and released as Selection-8. Like the selections 5 and 6, a majority of the plants of Selection-8 also manifest high resistance to rust in the field.

Among all the interspecific hybrids created in India, the one derived from the cross of diploid species C. liberica x C. eugenioides is the most interesting. This diploid interspecific hybrid was highly sterile to start with, but gave rise. to a sucker carrying the doubled number of chromosomes. The tetraploid sucker closely resembles certain Arabica types in morphology. Progenies from this sucker constituted tetraploid populations hat are highly resistant to leaf rust. This was released as Selection-II. Considering the sturdy and long lasting rust resistance manifested by the interspecific hybrids, crosses were effected between them and pure Arabicas to develop new selections. Thus, Devamachy hybrid was crossed to S.881 (Rume Sudan) derive lines S.2267, S.2268 and S.2269 and with S.333 to derive S.2139. These lines went on assessment for their performance in different locations. These are now exploited as Selection-5A and Selection-5B respectively. Both of them manifest high resistance to the disease and over 90% of the plants of these selections are also resistant to the disease in the field. HDT was crossed with several pure Arabicas and S~2790 derived from HDT x Tafarikela became popularly known as Selection-9 that is also highly resistant to leaf rust.

Dwarf mutants San Ramon, Caturra and Villa Sarchi were known to have been derived from Bourbon variety that was highly susceptible to leaf rust. In India, San Ramon was successively' crossed with S.1934, Cioccie/Agaro and HDT to incorporate as many genes as possible for rust resistance. Among the various lines released for commercial cultivation, Selection7.3 became popular in some locations on account of its high resistance to leaf rust.



Caturra was crossed with S.795 and Cioccie to derive hybrids carrying SH3 and SH4 genes respectively in combination with SH2, SH5 and SH5. A double cross hybrid was derived by crossing the two initial hybrids to combine the genes SH2,3,4,5 in one genotype. This genotype was created at CIFC, Portugal and was provided to CCRI for evaluation. This is commercially exploited as Selection-IO. Even though this selection manifests considerable resistance against leaf rust, the degree of resistance is not as high as in the S~lections 5, 6, 8 and 9. Another derivative of Caturra, a descendant of Caturra x HDT was also provided by CIFC in early 1980s and was released as Selection-12 named as Cauvery. Even though Cauvery manifested good resistance to rust initially, its resistance got diluted with successive generations.

Some of these selections are popular with the planters on account of their resistance, even though they are only moderate yielders.

Types and Sources of Resistance

The· knowledge of types of resistance such as vertical resistance (YR, generally conditioned by monoor oligo- genes) and horizontal resistance (HR, conditioned by polygenes) was gained in the recent times. Genes conditioning vertical resistance were reported to be epistatic to those conditioning horizontal resistance (epistatic= masking the expression). Ho'wever, our observations on various coffee selections indicate that there is a complementarity of these two types of resistance genes in coffee. When we look at our Selections from this perspective, it is discernible that SeIctions- SA, 6, 8, 9 and II possess the long lasting field resistance that is a combination of vertical and horizontal types of resistance. Thus, the VR-HR combinations of Selection-SA and Selection-9 manifested the longest lasting rust resistance in the field. Similarly, the longevity of resistance of Selection6 (artificial hybrid of Robusta x Arabica), Selection-8 (natural hybrid of Arabica x Robusta) and Selection-II (tetraploid derived from the hybrid of C. liberica x C. eugenioides) is indicative of these two types of resistance.

Newer Insights and Future Selection Protocols

Besides the knowledge on types and sources of resistance and complementarity of HR and· VR in coffee, insights were also gained into the biochemical manifestations of resistance and the molecular composition of chromosome regions closely associated with some resistance genes. This knowledge can be possibly exploited to identify disease resistant plants at an early stage of development in the nursery to select materials for the establishment of seed and wood gardens that, in turn, provide planting materials for supply to the growers.· Facilities for study and integration of biochemical and molecular markers in coffee breeding are being established at CCRI.

Gene pyramiding is another approach that is expected to buil.d long lasting resistance in the plant materials. In gene pyramiding, a selected genotype carrying dominant genes conferring the character of plant type (such as San Ramon, Caturra, Villa Sarchi in coffee) is crossed to a variety of materials carrying different resistance genes in order to derive a population in which all the plants appear physically similar but differ in their genetic constitution, however carrying all the known resistance genes. In the light of understanding that SH genes of coffee behave like Mendelian factors, it is possible to design crosses to brjng all known SH genes together. To defeat the resistance of such populations, the pathogen should evolve a race carrying many virulence genes. But, in nature, accumulation '~f virulence genes imposes a fitness' penalty on the pathogen, restricting its spread or sometimes eliminating it altogether. At CCRI, cross-bred populations were created by crossing Cauvery with a variety of Arabica and interspecific hybrid genotypes to evolve a gene pyramid. This is now being studied to isolate materials for commercial exploitation.

From the foregoing, it is evident that CCRI undertook the task of evolving rust resistant coffee strains when there was no knowledge of the nature of resistance in this plant, but contributed significantly by releasing the first strain within a time span of about 12 years and went on to evolve 12 strains with different degrees of resistance to be exploited commercially and is poised to develop more in times to come. Intellectually, the breeding team of the Institute contributed early knowledge on the segregation patterns and methods of evaluation of resistance and more recently on the biochemical and molecular aspects of resistance too. This rich background and experience with the plant and the pathogen will certainly lead to the evolution of newer coffee strains with stable and long lasting resistance, not only to the rust disease but also to a wide range of other pathogens and pests.