Responses of Compact Coffee Clones Against Coffee Berry and Coffee Leaf Rust Diseases in Tanzania

The utilization of resistant Arabica coffee (Coffea arabica) varieties is considered as the most economical control for coffee berry disease (CBD) and coffee leaf rust (CLR) in Tanzania. The resistance levels of varieties at field and laboratory conditions were assessed through their phenotypic disease reaction response to CBD and CLR. In this study sixteen (16) compact hybrids of C. arabica plus four (4) standard cultivars were evaluated under a range of environmental conditions in on-station and on-farm trials in Tanzania. Also four (4) Colletotrichum kahawae strains of the pathogen responsible for CBD infection; 2010/1, 2010/2, 2006/7 and 2006/14, and Hemileia vastatrix uredospores were used to test the sixteen (16) hybrids through artificial inoculation under controlled conditions (temperatures between 19 to 22 oC, R.H 100%). Results showed that a significant level of variability (P < 0.05) occurred between the sixteen (16) compacts, three (3) standard checks and N39 a commercial susceptible variety across trials. Compact genotype CVT14 (PNI086 x (N39 x Rume Sudan Selfed F2) showed resistance to the four strains of C. kahawae and Hemileia vastatrix. Differential reactions on compact genotypes were found to C. kahawae and H. vastatrix strains existing in different coffee growing regions in Tanzania; genotypes CVT4 (PNI088 x (SL34 x HdT) x Kent x Rume Sudan) and CVT13 (PRO127 x (Blue Mountain Jamaica x Cioccie) x Rume Sudan) showed partial resistance to C. kahawae strains 2010/1, the genotypes were susceptible to strains 2010/2 but resistant to strains 2006/7 and 2006/14. This shows that host response reaction can be used as criteria for varietal assessment when evaluated at different locations.


Introduction
Coffee production in Tanzania is still constrained to a large extent by the ability of the growers to control coffee berry disease (CBD) and coffee leaf rust (CLR).These are fungal diseases; CBD is caused by Colletotrichum Kahawae Waller and Bridge Sp.Nov., and CLR by Hemileia vastatrix Berk et Br.Coffee berry disease attacks all stages of the developing crop from pinhead to ripe fruits (Mulinge, 1970), which may destroy 30 to 90% of the crop if the weather conditions are favourable to its epidemics (Ngulu et al., 1998).Coffee leaf rust disease infection can cause severe leaf defoliation leading to die-back of primary branches, followed by death of the coffee tree.
The traditional Arabica coffee varieties commercially grown in Tanzania are all susceptible to CBD and CLR.This is despite their fine quality coffees, in high demand by buyers and roasters (Wrigley, 1989).The varieties including N39, KP423, KP162 and H66 among others occupy over 80% of the total Tanzanian coffee acreage (Fernie, 1970;Robinson, 1964).They are protected against CBD and CLR by an intensive fungicide spray programme that accounts for up to 50% of the production costs (Gustave et al., 2007;Teri et al., 2004;Omondi et al., 2000).The fungicides are not only uneconomical but also environmentally unsafe.
It is for this reason that selection for resistant coffee types was initiated in Tanzania in stabilizing yields and quality of coffee.Coffee Research at Tanzania Coffee Research Institute (TaCRI) initiated a programme to develop compact hybrids in 2003/04 involving Catimors, Hibrido de Timor (HdT), Rume Sudan and varieties contributing attributes of good quality coffee; N39, KP423 and SL28.Most of Catimor lines carry the dominant resistance S H 6 and S H 9 introgressed from HdT (Betterncourt et al., 1992).These two genes confer resistance to H. vastatrix (Rodrigues Jr. et al., 2000).Studies by van der Vossen and Walyaro (1980) revealed that HdT possesses a gene for CBD resistance on the T-locus, and that dominant R and recessive k genes found in Rume Sudan contribute to CBD resistance in the cross of Coffea arabica.The compact hybrid breeding lines developed were established in four on-station and six on-farm trials to study genetic responses under varying environmental conditions in Tanzania.Plant pathogen distribution and variability forms important part in assessing the response of crop varieties under diverse conditions (Brenda, 2011;Gichuru & Omondi, 2010).The aim of this work was to identify compact coffee clones that are resistant to CBD and CLR in specific or across ecosystems.

Description of the Study Area
The trial sites were located in low altitude coffee growing areas prone to CLRD (1000 to 1200 m. a.s.l), at medium altitude where CLR and CBD prevail (1200 to 1400 m. a.s.l) and high altitude areas where there is high incidence of CBD (> 1400 m. a.s.l).These sites have considerable variation in pathotypes for CBD and CLR.The study areas are shown in Figure 1.

Coffee Genotypes
Sixteen compact coffee genotypes, and four check varieties (coffee hybrid SC9, SC13, Catimor PNI088 and N39), were tested for host x pathogen interaction across agro-ecological areas (Table 2).Hybrid seeds of the same breeding lines were produced, raised to the hypocotyl stage (5-6 weeks old, 5-6 cm tall) and tested for genotype x pathogen interaction to CBD and CLR under controlled conditions at temperature of between 19-21 ºC, and R.H. of approximately 100%.Note: Ctr = Catimor line.

Collection and Preparation of Colletotrichum kahawae Inoculum
Green coffee berries infected with CBD at the black lesion stage were collected close to trial sites and preserved in a paper bag.Inside the paper bag, green infected berry samples were sandwich between layers of newspaper and kept in a cool box at 18 to 20 ºC for the pathogen to be viable for successful subsequent isolation (Johnston & Booth, 1983).Colletotrichum kahawae was isolated from the survey samples using the method described by Beynon et al. (1995).Infected green coffee berries with CBD lesions were surface sterilized by immersing for 15 min in a 5% solution of sodium hypochlorite (available chlorine 0.5% W/V) and rinsed in three changes of sterile distilled water.Infected coffee berry tissues were then removed aseptically and incubated on distilled water agar at 22 ºC.Hyphae from advancing edges of the tentative C. kahawae colonies were then transferred to Malt Extract Agar (MEA) (Beynon et al., 1995).The cultures were incubated at 22 ºC for 10 days to allow colony growth.Preliminary confirmatory tests of colony texture of C. kahawae isolates on MEA were based on mycological colour chart developed by Rayner (1970).Usually C. kahawae colonies have grayish colony texture.
The pathogenicity of C. kahawae was tested on 20 fully expanded, soft green berries ( 14 weeks old from date of flowering) of the susceptible variety N39.Coffee berries were surface sterilized, placed on damp sterilized sand in plastic boxes and inoculated with a 0.02 ml drop of the standard C. kahawae spore suspension at 2.0 x 10 6 conidia per ml using a pipette.The inoculated green coffee berries were incubated at 25 ºC and observed for CBD symptom development for 14 days (Figure 2).Days to first symptoms appearance and percentages of infected berries were recorded.

Collection and Preparation of Hemileia vastatrix Inoculum
Leaves of N39 with fresh rust lesions of Hemileia vastatrix were collected uredospores scraped into conical flask contained sterilized dH 2 0. Rust uredospores were collected from live plants because Hemileia vastatrix is an obligate parasite and it is impossible to grow and multiply through culture media (Kushalappa & Eskes, 1989).
The solution was then shaken to allow dispersion of uredospores, and to allow uniform dispersion, a drop of tween 80 is added into a solution and then shaken.Then uredospores concentration calibrated at 1 x 10 6 /ml.

Field Experiments
Multi-location trials located at Lyamungu (1268 m a.s.l),Ugano (1562 m a.s.l),Mbimba (1605 m a.s.l) and Mwayaya (1530 m a.s.l) were used to identify compact coffee clones that are resistant to CBD and CLR pathogens in specific or across ecosystems.
The data were collected in the already existing compact trees that were used to design experimental plots.A completely randomized block design (CRBD) with three replications experimental design was used at Lyamungu, Mbimba and Ugano in the study.Each plot consisted of 24 trees but only the inner 8 trees were considered for precise data collection.The spacing between coffee trees was 2.0 m and within 1.5 m.A set of two branches having five to six berry clusters per branch (about 40 berries per cluster) of green expanding coffee berries at 8 to 12 weeks old, were used for CBD assessment.Disease incidence was done by counting the number of CBD coffee infected berries against uninfected berries, and then the percentage berry infection was computed.
number of CBD coffee infected berries in berry clusters Disease incidence 100 total number of berries in berry clusters   For CLR the same trees were used as a reading unit to evaluate disease reaction of the genotypes using 0-9 scale by Eskes and Toma-Braghini (1981); whereby 0 describes absence of lesions and 9 describes intense lesions associated with leaf shedding.

Experiments Under Controlled Conditions
Colletotrichum kahawae strains originated from four (4) ecosystems; Moshi, Kigoma, Mbeya and Mbinga were used to test the level of resistance of sixteen ( 16) Coffea arabica compacts.A set of two branches having five to six berry clusters per branch of green expanding berries were kept in a conical flask having sterilized dH 2 0 to sustain their vigour.These were kept in sterile controlled conditions at temperature of 19-21 ºC, and R.H. of approximately 100%.The berries were artificially inoculated with strains of C. kahawae suspension at 2 x 10 6 conidia/ml.The quantity of the conidia concentration was determined and standardized using haemocytometer.
The C. kahawae strains were 2010/1 (Kigoma), 2010/2 (Lunji-Mbeya), 2006/7 (Mbinga) and 2006/14 (Kibosho-Kombo Moshi).One block representing check was sprayed with dH 2 0 only.The mean percentages of green coffee berries infected with CBD were recorded at 10, 14 and 18 days after artificial inoculation.Similarly F 1 hybrid seed produced through artificial pollination between male parents (Table 2) and the selected compact lines PNI086, PNI088 and PRO127 were used to determine genotype x CBD pathogens interaction using the four C. kahawae strains.F 1 of each of the 16 hybrids represented by 40 hypocotyls were raised in a plastic box at a spacing of 5 cm x 5 cm containing sterile sand (Figure 3).The experiment was arranged in the laboratory in a completely randomized design (CRD) with three replications.Hypocotyls at 5-6 weeks or 5-6 cm were sprayed and inoculated with suspension of C. kahawae strains at 2.0 x 10 6 conidia/ml twice at 48 hr intervals using the method by van der Vossen et al. (1976).Three weeks after the date of first inoculation, coffee seedlings were individually scored for CBD symptoms which developed on the hypocotyls using a scale with a range of 0-4 which was developed by van der Graff (1982), where 0 is nil and 4 intense CBD infections.For each genotype Disease Index Reaction for sixteen (16) F 1 hybrids was determined as follows: Where, i is the disease class, ni is the number of seedlings in class i and n is the total number of seedlings scored.Leaves of the terminal node still tender and succulent arranged incompletely randomized (CRD) were used to evaluate the levels of resistance of the compact genotypes using coffee rust races collected from N39 from the four coffee ecosystems.Ten (10) leaves of each of the genotype kept in a sterile box were inoculated by spreading dry coffee rust uredospores on the lower surface of the leaves using sterilized camel's hair brush as described by d 'Oliveira (1965).After inoculation the leaves were sprayed with dH 2 0 and then placed in moist chamber for 48 hr to allow germination of spores and infection.The assessment of the reaction on the leaves was done using a quantitative scale developed by d' Oliveira (1965); where i is without infection and X highly variable size of sporulating pustules.The assessment on coffee rust disease reaction was concluded 45 days after the artificial inoculation, the final assessment was considered to confirm the resistance of the genotypes.
The data were analyzed using Genstat software, and mean separation was done using Tukey's multiple tests.Data for CBD and CLR were not transformed as the raw data were symmetrically distributed; between the 16 compact coffee genotypes and N39.Key: 0 = absence of CBD symptoms; 1 = few (1-2) and small chlorotic lesions; 2 = more than 2 brownish lesions or coalescence brownish lesion; 3 = large brownish lesions with abundant black dots and or black lesions; 4 = dead hypocotyls.*Assessment 14 days after artificial inoculation of C. kahawae isolates.

Reaction of Compact Coffee Genotypes to CBD and CLRD Under Various Ecosystems Under Natural Conditions
Figures 4-5 summarizes the results on reaction of compact genotypes to CBD and CLR for on-station trials under various ecosystems under natural disease infection.Figure 4 show results on percentage CBD infection of the compact genotypes and check varieties.High levels of CBD infection were noted in N39 across the sites, but there was nil infection in the sixteen compact genotypes, SC9, SC13 and Ctr088.Error bars with 0.05 values show significance differences between the sixteen compact genotypes, SC9, SC13, Ctr088 and N39.
Results in Figure 5 show that, N39 had high CLR reaction compared to the rest of the genotypes.There is also significance difference shown by error bars with 0.05 values.
Table 4 shows reaction of compact genotypes to CBD and CLR for on-farm trials across coffee ecosystems under natural disease infection.Results show that across six locations in farmer's plots, the only cultivar which was susceptible to CBD and CLR was N39.The sixteen compact genotypes showed resistance to CBD and CLR under natural conditions.Analysis show high interaction of CBD effect on season, location and genotypes, but for CLR only for location and genotypes (Table 5).Key: Percent infection: 0-5 = Resistance (R); 6-20 = Partial Resistance (PR); 21-39 = Moderately Susceptible (MS); 40-100 = Highly Susceptible (HS) (Fernie, 1963).Means followed by a common letter within a column do not differ significantly (P < 0.05).Table 8 presents the results on the reaction of C. kahawae strains on hypocotyls of the compact genotypes.The results show a significant (P < 0.05) level of variability between the sixteen (16) compact genotypes, three checks and N39.Six genotypes of compact hybrids had DIR scores between 0-25 viz.; CVT2, CVT4, CVT5, CVT7, CVT8 and CVT13.Among the checks, Catimor PNI088 showed outstanding performance in terms of CBD resistance.Key: DIR 0 -25 = Resistant, 26 -50 = Moderately Resistant; 51 -75 = Moderately Susceptible and 76 -100 = Susceptible.Means followed by a common letter within a column do not differ significantly according to Tukey's (P ≤ 0.05).

Reaction of Compact Coffee Clones to CLR Under Controlled Conditions
Table 9 presents the results on the reactions of compact coffee leaves upon artificial inoculation of H. vastatrix uredospores collected from N39.Out of sixteen ( 16), ten (10) compact genotypes showed immune reaction from the rust inoculum collected from N39 and KP423 variety.N39 was the most susceptible genotype to CLR.Key: N39 is susceptible to all available rust races.

Discussion
The presence of appropriate plant organ, conducive weather conditions and a pathogen helps in assessing the response of genotype for its resistance, partial resistance, moderate susceptible and susceptibility.From flowering, which occurs shortly after the end of the dry season, coffee berry takes about nine months to maturity (harvest).During the first four weeks, the berry does not increase in size; it remains at the "pinhead" stage which is normally resistant to CBD (Mulinge, 1970).The expanding berry (4-16 weeks after flowering) is the most susceptible stage; this is unlike fully expanded green berries, which are resistant (Mulinge, 1970).The presence of free water (rain, mist or dew) on berry surfaces and temperatures between 21 and 23 ºC are favourable condition for infection and development of epidemics (Nutman & Roberts, 1960;Bock, 1956).Incubation period varies between 3 days with soft green berries (Bock, 1956) to 3 weeks with hard green berries (Mulinge, 1970).Conidia are only released and dispersed by rain between and within coffee trees.
Epidemiology for CLR requires optimum temperatures of between 21 and 25 ºC for the development of H. vastatrix (Guggenheim & Harr, 1978).Rain is an important vector for short range distribution of the uredospores, creating suitable conditions for germination (Rayner, 1961).When these conditions prevail across sites they may allow infection and thereby a separation between susceptible and resistant genotypes among the compact coffee.
Rainfall and temperature conditions at Lyamungu for 2009/10 were conducive for the development and infection of CBD and CLR (Figures 6 and 7), N39 the susceptible variety succumbed to CBD and CLR, but compact genotypes were resistant.Although van der Vossen et al. (1976) found possible correlation between hypocotyls test and reaction of CBD on expanding berries was not the case in this study.Genotypes CVT10 (Ctr127 x (N39 x Rume Sudan selfed F 2 ) and CVT14 (Ctr086 x (N39 x Rume Sudan Selfed F 2 ) showed CBD resistance when green expanding berries were artificially inoculated, but the hypocotyls did succumb to the disease.As Gustavo et al. (2007) (1982) and Omondi et al. (2001) which also revealed variations on the effect of the pathogen and resistance of the coffee genotypes.
Studies done previously on C. arabica genotypes variety x C. kahawae strains reported existence of interaction, but found to be less significant to suggest conclusively that differential interaction exists (Omondi et al., 2000).
Further studies within C. kahawae at the DNA level have also revealed limited polymorphism (Beynon et al., 1995).The formation of hyphal fusions during infection of C. kahawae is a likely mechanism creating pathogen variations as there is involvement of an exchange of genetic materials (Mwang'ombe et al., 1992).
Out of sixteen ( 16), nine compact genotypes showed immune reaction from the coffee rust inoculum collected from N39 and KP423.Immune reaction may possibly be resulted from HdT and Catimors which provides genes for resistance to CLR.HdT has been utilized in developing resistant coffee hybrids to CLR in Kenya (Gichimu, 2012).Variety N39 was the most susceptible genotype to CLR.The pathogen causing CLR, H. vastatrix Berk et Br., breaks the coffee resistance genes of the most of the cultivars.This is because the physiologic races of the pathogen causing CLR are increasing.Varzea et al. (2002b;2002c) reported 40 physiological races.In Tanzania, Rodrigues Jr. et al. (1975) reported the existence of seven races but by 2007, additional coffee rust races were reported (TaCRI, 2009;CIFC, 2007).The available races might form part of the CLR inoculum collected from N39 used to test the compact genotypes.While N39 is reported to be attacked by all possible coffee rust races, KP423 is attacked by the most virulent coffee rust race I (Kushalapa & Eskes, 1989).In spite this situation, the host resistance of the compact genotypes which has genes S H 6 and S H 9 provide adequate protection against CLR epidemics (van der Vossen, 2005;Rodrigues Jr. et al., 1975).It can therefore be deduced that the nine identified compact genotypes showed resistance to CLR

Conclusion
In this study the results indicate that there is a significant magnitude of resistance of the compact genotypes when tested across coffee growing regions and under controlled conditions.However, there are compact genotypes which are resistant to CBD and CLR and therefore, can be adapted across the prevailing coffee disease strains.These coffee genotypes are potential sources of resistance to CBD and CLR in breeding programmes for developing future commercial coffee cultivars.This study also indicates that C. kahawae strains are likely to have adapted to the coffee genotypes.However, further investigations are needed to establish the possibility of the existence of physiologic races of the C. kahawae strains.

Figure 1 .
Figure 1.Map of Tanzania showing study areas highlighted with red dotsSpecific description of the study locations in terms of altitude, weather conditions (annual precipitation, humidity, temperatures) and soil types are summarized in Table1.

Figure 2 .
Figure 2. Pathogenicity tests of Colletotrichum kahawae isolates on green berries of N39

Figure 4 .
Figure 4. Reaction of the coffee genotypes to coffee berry disease across ecosystems

Table 2 .
Coffee genotypes of compact breeding lines

Table 4 .
Reaction of compact coffee genotypes to Colletotrichum kahawae and Hemileia vastatrix evaluated under natural conditions across six on-farms in different agro ecological zones for 2010/11

Table 5 .
ANOVA summary to determine effect of season, location and coffee genotypes with CBD and CLR *** Highly significance differences.

Table 6 .
Reaction of green berries of compact genotypes 18 days after artificial inoculation to Colletotrichum kahawae strains

Table 7 .
ANOVA summary to determine genotype interaction with Colletotrichum kahawae strains using green berries of compact genotypes

Table 8 .
Disease Index Reaction of the compact genotypes at the hypocotyls stage to four Colletotrichum kahawae strains

Table 9 .
Reaction of coffee rust disease on Coffea arabica compacts 42 days after artificial inoculation of uredospores collected from N39 and KP423 document, different resistance in reactions was found in hypocotyls test in relation with fruits in coffee inoculated with Colletotrichum gloeosporioides isolates.Van der Graff (1982)observed different resistance reaction to C. kahawae in the types of CBD evaluation methods.The coffee genotypes CVT2 (Ctr088 x (SL34 x HdT) x Kent x Rume Sudan), CVT4 (Ctr088 x (SL34 x HdT) x Rume Sudan), CVT5 (Ctr088 x (Rume Sudan x Catuai), CVT7 (Ctr088 x (HdT x N39) x SL28) x (N39 x Rume Sudan)), CVT8 (Ctr 088 x (N39 x HdT) x (N39 x HdT) x Rume Sudan) and CVT13 (Ctr127 x (Blue Mountain Jamaica x Cioccie) x Rume Sudan) showed resistance to the four C. kahawae strains used in the study.Variety N39 showed high susceptibility to the four C. kahawae strains when CBD hypocotyl test was applied.The overall observation here is that expression of resistance varied with the test C. kahawae strains and the hybrid crosses.These results are in agreement with the findings of van der Graff