Raw coffee processing yield affected more by cultivar than by harvest date 12

Transcription

1 Raw coffee processng yeld affected more by cultvar than by harvest date 12 Teodoro Espnosa-Solares, 3 Juan Gullermo Cruz-Castllo, 4 sval Antono Montesnos-López 3 and Arturo Hernández-Montes 3 J. Agre. Unv. P.R. 89(3-4): (2005) ABSTRACT Converson of cherry and parchment coffee (Coffea arábca L.) to green coffee was determned durng two consecutve years of harvest for eght dfferent coffee cultvars grown n Mexco. 'Mass balance' was useful to adjust the frut humdty to 0.12 g/g for coffee bean raw processng. Unvarate and multvarate analyses showed sgnfcant dfferences (P < 0.05) among the cultvars. A lnear dscrmnant functon based on twelve varables measured on several dates n two consecutve harvest years showed that 'Typca' and 'Mundo Novo' had hgh effcences n conversons from cherry to green coffee on three harvest dates n 1999 and one n In contrast, 'Bourbon' and 'Yellow Caturra' requred greater amount of fresh frut to obtan 46 kg (1 quntal) of green coffee. The cultvar effects on the parchment to green coffee bean yeld were not clearly dentfed. Harvest dates studed dd not sgnfcantly nfluence the converson of cherry or parchment to green coffee bean yelds n the two years evaluated. Key words: Coffea arábca, coffee beans, raw coffee processng, parchment coffee, green coffee RESUMEN Influenca del cultvar y de la fecha de cosecha en el benefcado ndustral de granos de café Se estudaron conversones de café cereza y café pergamno a granos de café verde o plado (Coffea arábca L.) en ocho cultvares de cafeto, durante dos años consecutvos de cosecha en Méxco. Un 'balance de masas' fue útl en ajusfar la humedad del fruto a 0.12 g/g para granos de café a procesar. Análss unvarados y multvarados mostraron dferencas sgnfcatvas (P < 0.05) entre los cultvares. Una funcón lneal dscrmnante basada 'Manuscrpt submtted to Edtoral Board 9 March The authors thank the Unversdad Autónoma Chapngo for fundng ths study. M.C. Esteban Escamlla, Ing. Francsco López, Ing. Amlkar López, Ing. Lauro Rayón and Ing. Rutlo Enríquez are thanked for ther help on the coffee plantaton and for supplyng techncal nformaton. Also Dr. Bruce R. MacKay, from the Insttute of Food Nutrton and Human Health (Massey Unversty, New Zealand), and Dr. Carlos Guadarrama-Zugast from the Center for the Development of Coffee Regons (Unversdad Autónoma Chapngo, Méxco) are thanked for ther manuscrpt comments. 3 Departamento de Ingenería Agrondustral. Unversdad Autónoma Chapngo. P.. Box 161, Chapngo, 56230, Edo. de Méxco, Méxco. E-mal: espnosa@chapngo.mx 4 Centro Regonal Unverstaro rente. Unversdad Autónoma Chapngo. P.. Box 49, Huatusco, Veracruz, Méxco E-mal: jcruzcastllo@yahoo.com 169

2 170 ESPINSA-SLARES ET AL./CFFEE PRCESSING YIELD en doce varables meddas en varas fechas durante dos años consecutvos de cosecha mostró que 'Typca' y 'Mundo Novo' tuveron una alta efcenca en conversones de café cereza a granos de café verde en tres épocas de cosecha en 1999 y una en el En contraste, 'Bourbon' y 'Caturra Amarllo' requreron una mayor cantdad de fruto fresco o café cereza para obtener 46 kg (1 Qq) de granos de café verde. La nfluenca de los cultvares en el rendmento ndustral de café pergamno a granos de café verde no se dentfcó claramente. Las fechas de corte estudadas no nfluyeron sgnfcatvamente en la conversón de café cereza o café pergamno a granos de café verde en dos años de estudo. Palabras clave: Coffea arábca, granos de café, procesamento ndustral de café, café pergamno, café verde, café plado INTRDUCTIN Coffee s the most mportant tropcal crop consumed worldwde. Accordng to van der Stegen (2003), the world producton of coffee amounts to ~110 mllon 60-kg bags per year (6.6 mllon tons/year), wth about 90 mllon bags n the exportaton market. The adequate mosture content for storage of green coffee beans s near 0.12 g/g (Atmawnata, 1995); consequently, the reference for trade operatons s based on ths mosture value. The nternatonal coffee market recognzes factors that nfluence coffee qualty: speces, plantaton alttude, lattude, cultvars, raw coffee processng type, physcal qualty of green coffee and sensory propertes of coffee beverage (Vlla, 1990; Puerta- Quntero, 2000a; Fgueroa-Solares et al., 2000; Njoroge, 1998; Reyna, 1989; González-Arcos, 1996). Coffee beverages are made from roasted green coffee beans of two speces wth economc mportance, Coffea arábca and C canephora (P.), known n the trade market as arábca and robusta, respectvely (Martín et al., 2001). Coffea arábca presents hgh standards n aroma and flavor whereas C canephora contans hgher caffene and soluble sold levels (Cléves-Serrano, 1998; Ky et al., 2001). The transformaton from coffee frut cherres to green coffee beans, known as raw coffee processng, plays an mportant role n the fnal qualty of the beverage. In fact, coffee qualty s obtaned n the feld and t s preserved durng raw coffee processng. Wet and dry processes are the two man methods used for green coffee producton. Wet processng dffers from dry processng bascally n the producton of parchment coffee as an ntermedate product, whereas dry processng has dred cherres as an n-between product (Fgure 1). Prevous to trade, parchment coffee can be stored for almost one year wthout any loss because of the solaton provded by the endocarp. Mexcan green coffee beans are manly obtaned from C arábca and produced by wet processng. The beverages produced through ths process have a hgh acdty and dstnctve aroma. Frut yeld, tolerance to pathogens, and envronmental adaptaton have been key crtera for coffee breedng (Vllaseñor, 1987). However,

3 J. Agre. Unv. P.R. VL. 89, N. 3-4, JULY-CTBER Cherry coffee Wet process Pulpng 1 Fermentng or mechancal muclage remoton Washng Dryng Parchment coffee Hullng I Sze, shape, densty and color classfcaton 1 Washed green coffee beans Dry process ]\ Dryng 1 Dryed cheny coffee Hullng 1 Sze, shape, densty and color classfcaton 1 Natural green coffee beans FIGURE 1. Wet and dry cherry coffee processng. snce green coffee s the fnal product, t s mportant to consder wet processng yelds n breedng programs. Specfcally, t s valuable to study the converson from cherry to parchment and from parchment to green coffee beans. In commercal coffee trade agreements, 57.5 kg of parchment coffee s consdered as one 'Quntal' (Qq), whch corresponds to 46 kg of green coffee at 0.12 g/g of humdty. Some authors have demonstrated that raw coffee processng yeld s affected by cultvar, plant age, locaton and frut rpeness grade (Castllo, 1981; Guyot et al., 1996; Montagnon et al., 2000; Soto-Pnto et al., 2000). Coffee growers assume that

4 172 ESPINSA-SLARES ET AL./CFFEE PRCESSING YIELD processng yeld ncreases as harvest date advances. In Afrca, there s nformaton on ths assumpton (Haarer, 1980). However, the nfluence of harvest date on raw coffee processng has not been carefully studed n Mexco. In fact, there are only a few local reports concernng ths ssue. Puerta-Quntero (2000b) reported the negatve nfluence of mmature beans on wet processng. Parchment coffee yeld of mmature beans was reduced by 7% as compared to that of rpe beans. Ths author also showed that mmature beans affect physcal and sensory propertes; thus, the hgher percentage of rpe coffee n a sample, the hgher number of good qualty cups s obtaned. Reyna (1989) found that cultvar nfluenced processng yeld. In that study, 'Typca' requred kg of rpe cherry frut to produce 1 Qq of green coffee, followed by 'Vlla Sarchf (241.5 kg), 'Catuaf (246.5 kg), 'Pacas' (250.2 kg), and 'Caturra' (256.2 kg). Guadarrama and Trujllo (2001) showed that the converson from cherry to parchment n 'Mundo Novo' was nfluenced by the envronmental condtons of the plantaton. The present work was conducted to determne the nfluence of harvest dates on green coffee yelds for several coffee cultvars n a Mexcan tropcal upland envronment producng hgh qualty green coffee beans. Mass balance, unvarate and multvarate analyss of varance, and canoncal dscrmnant analyss were used to evaluate converson changes from cherry to green coffee and parchment to green coffee obtaned by wet processng n two consecutve years of harvest. MATERIALS AND METHDS Plantaton and harvest. An 18-year-old coffee (C. arábca) plantaton establshed on an Andosol sol, located at 1,344 m above sea level at the Centro Regonal Unverstaro rente of the Unversdad Autónoma Chapngo n Huatusco, Veracruz, Mexco (19 09' N and 96 57' W), was used n the experment. The study area had a mean annual ranfall of 1,700 mm, and an annual temperature average of 17.2 C. Eght cultvars wdely grown on local plantatons were selected: Garnca, Typca, Pluma Hdalgo, Catua, Red Caturra, Yellow Caturra, Mundo Novo, and Bourbon. Accordng to the classfcaton of type of canopy (Gobb, 2000), the producton system had a 'technfed' shade, wth predomnancy oí Inga sp. trees. Frut harvested n 1999 and 2000 was consdered n the study. Three hand harvests were performed each year. The frst two harvests were n January and February n both years, and the last n March of 1999, and May of The crteron adopted n defnng the harvest date was the red color of the frut skn as ndcator of rpenng. Yellow Caturra was harvested when the frut attaned a yellow color.

5 J. Agre. Unv. P.R. VL. 89, N. 3-4, JULY-CTBER Raw coffee processng. Three replcates of 2-kg samples of cherry coffee were randomly harvested and processed by the wet method for each cultvar and harvest date. nly rpe fruts were used throughout all the experments. The frut processng ncluded pulpng, fermentng, washng, and dryng. The product obtaned was parchment coffee at a humdty of about 0.12 g/g. Afterwards, parchment was hulled to obtan green coffee beans. Mass balance and converson factors. Mass balances were used to standardze the samples to 0.12 g/g of humdty (Hmmelblau, 1997). Cherry to green coffee (CF Ch. G ) and parchment to green coffee (CF P. G ) converson factors were used for comparng the raw coffee processng yeld. Both converson factors were defned by usng the mass of the samples of each treatment and defned by equatons 1 and 2. These equatons defne, respectvely, the amount of cherry and parchment needed to obtan 46 kg (1 Qq) of green coffee beans. 46 kg ^P _ mass of green coffee (1) Ch mass of cherry 46 kg p, _ mass of green coffee (2) P mass of parchment Statstcal analyss. Unvarate (Steel and Torre, 1980) and multvarate analyss of varance (Krzanowsk, 1988) and canoncal dscrmnant analyss (Cruz-Castllo et al., 1994) were performed to compare the nfluence of harvest date and cultvar on raw coffee processng yeld, consderng the cherry to green coffee and parchment to green coffee converson factors over two-year harvests. The SAS (1989) ANVA and CANDISC procedures were used to perform these statstcal technques. RESULTS AND DISCUSSIN Mass balance was useful n obtanng the equvalent amount of parchment and green coffee beans at the typcal commercal humdty (0.12 g/g). Cherry to green coffee yelds were sgnfcantly (P < 0.05) affected by the cultvars on most of the dfferent harvest dates (Table 1). The Typca cultvar acheved the hghest cherry to green coffee converson, to complete kg of green coffee (1 Qq) n March 1999 and n May 2000 wth and kg, respectvely. Coffee growers n Huatusco, Mexco, clam that the best yeld conversons from cherry to

6 -J > ^ TABLE 1. Harvest dates n January (J), February (F), March (M), May (My) and factor converson means (kgj 1 of cherry to green (CF chg ) and parchment tc < green coff ee(cf p Jf or eght dfferent coff: ee cultvars n the harvest seasons of 1999 and Cultvars Bourbon Catua Garnca Mundo Novo Pluma Hdalgo Typca Red Caturra Yellow Caturra Means and SE per harvest date J- CF Ch-G a ab a ab a b a a ±3.4 J. CF P. G 58.0 a 56.4 be 57.7 ab 56.5 be 57.5 abc 56.2 c 57.7 ab 57.6 ab 57.2 ± Season F- CF Ch-G ab ab ab b b b a ab ±2.7 E CF P. G 56.1a 56.4 a 56.8 a 56.2 a 57.6 a 55.9 a 57.3 a 56.1a 56.5 ±0.1 Mean separaton n each column by Tukey at 5%. M. CF ch. G a abc ab be abc c ab ab ±2.9 M. CF P. G 56.3 ab 55.8 ab 55.9 ab 56.0 ab 56.3 ab 55.5 b 56.1 ab 56.9 a 56.1 ±0.1 J- CF Ch-G a ab ab ab ab b ab ab ±5.0 J. CF P. G 55.7 a 54.6 a 54.7 a 55.0 a 53.9 a 53.0 a 54.3 a 53.5 a 54.3 ± Season F- CF Ch-G a a a a a a a a ±2.0 E CF P. G 56.1a 55.3 a 56.2 a 56.0 a 57.5 a 57.8 a 58.9 a 56.4 a 56.8 ± 0.3 My CF ch. G My. CF P. G a ab ab b ab c ab ab ± a 54.0 a 54.6 a 54.2 a 54.2 a 53.4 a 55.9 a 54.6 a 54.4 ± 0.2 H > o > H > tr 1 H Cf ü

7 J. Agre. Unv. P.R. VL. 89, N. 3-4, JULY-CTBER green coffee occurred at the end of the harvest perod. In the present work, there was not any late hgh harvest ndustral yeld trend when all cultvars were consdered. For example, the frst harvest n January 2000 acheved a hgh cherry to green coffee yeld converson across all the cultvars (236.1 kg). Lkewse, ths converson was hgh n February durng the frst harvest season (230.9 kg). Partcular cultvar effects nfluenced by harvest date n ths regard were shown by Bourbon, Catuaí, Mundo Novo, Typca, and Yellow Caturra wth hgh cherry to green conversons n January for each season (Table 1). In contrast, Red Caturra, Pluma Hdalgo, and Garnca dd not follow any harvest trend n the cherry to green coffee converson n ether season (Table 1). Thus, pre-harvest factors such as management of the plantaton shade (Muschler, 2000) may be more mportant n some cultvars for achevng adequate yeld converson values rrespectve of the tme of harvest. In a study n Costa Rca, cultvars wth longer perods of frut growth on the plant produce the heavest green coffee beans (Alpzar, 2000), but there are no smlar reports for Mexco. Typca was sgnfcantly dfferent from the other cultvars n May 2000, achevng kg of cherry for 46 kg of green coffee (Table 1). The parchment to green yeld converson n the frst season was sgnfcantly dfferent among the cultvars n January and March for Catuaí, Mundo Novo, Pluma Hdalgo, and Typca, requrng less parchment coffee to attan 46 kg of green coffee (Table 1). In the second season, all cultvars had smlar parchment to green coffee conversons (Table 1). The multvarate analyss of varance showed sgnfcant dfferences (P < 0.05) among the eght dfferent cultvars, where cherry to green and parchment to green coffee yeld converson factors were measured on dfferent harvest dates, n two consecutve harvest seasons. The canoncal dscrmnant analyss showed that the frst two canoncal functons were sgnfcantly dfferent (P < 0.05) from the other four canoncal functons, thus ndcatng that the eght dfferent cultvars dffered most n these two lnear functons, whch accounted for 91.2% of the varaton (Table 2). The frst canoncal dscrmnant (CDFj) functon explaned 83.2% of the varablty, and t was hghly assocated wth the cherry to green coffee conversons n March 1999 and May 2000, accordng to the correlaton coeffcents between the CDFj and all the converson values. The standardzed canoncal coeffcents n the CDFj showed large absolute values for the converson of cherry to green coffee n May 2000, and January, February, and March Absolute values were also hgh for the parchment to green coffee conversons n May 2000 and February Typca, wth the hghest absolute score (Table 3), was dstnctly dfferent from the other cultvars. Wth Typca, less harvest yeld was needed to obtan 46 kg of

8 176 ESPINSA-SLARES ET AL./CFFEE PRCESSING YIELD TABLE 2. Standarzed canoncal coeffcents (SCC) and correlaton coeffcents between canoncal dscrmnant functons (CDF p CDF^, and factor conversons of cherry to green (CF ch _g), and parchment to green coffee (CF P _Q) for eght dfferent coffee cultvars n the harvest seasons of 1999 and CDF CDF 2 Varables Equaton 3 SCC T2 SCC T Season January. CF ch. G January. CF P. G February. CF ch. G February. CF P. G March. CF ch. G March. CF P. G Season January. CF ch. G January. CF P. G February. CF ch. G February. CF P G May. CF ch. G May. CF P. G ^ Xs x 3 x 4 x 6 Xs X, x 8 X«Xo X n X Varance explaned (%) green coffee n 1999 [219 kg (Feb); 236 kg (Jan); and 242 (Mar)]; and n 2000 [200 kg (Jan); 231 kg (May)] (Table 1). Reyna (1989) also demonstrated that Typca frut was hghly effcent at producng green coffee. Regardng parchment to green coffee converson for Typca, the low correlaton coeffcent values (Table 2), and non sgnfcant (P < 0.05) mean values (Table 1) n February 1999 and May 2000 ndcated no dstnct TABLE 3. Means of standardzed canoncal scores of the frst two canoncal dscrmnant functons (CDFs) for eght dfferent coffee cultvars n the harvest coffee seasons of 1999 and Coffee Cultvars CDF CDF, Bourbon Yellow Caturra Red Caturra Catua Pluma Hdalgo Garnca Mundo Novo Typca 13.3 a 7.8 b 2.7 c 1.1c 0.7 cd -1.7 d -5.5 e f Mean separaton n each column by Tukey at 5%. 3.1a -2.7 cd 1.4 ab -0.1 be -4.8 d 3.5 a -1.6 c 1.2 ab

9 J. Agre. Unv. P.R. VL. 89, N. 3-4, JULY-CTBER advantages wth respect to the other cultvars n completng 46 kg of green coffee. These fndngs could be attrbuted to the dependence of ths converson on the endosperm-endocarp rato, whch vares wth the natural dfferences among fruts (Puerta-Quntero, 2000 b). Although Mundo Novo was sgnfcantly dfferent from Typca (Table 3), t also scored hgh n ths lnear functon and showed effcency n attanng 1 Qq on the above mentoned dates. Typca presented hgher frut weght and larger seeds but lower total yeld on the coffee plantatons n comparson wth the other cultvars studed (Vllaseñor, 1987). In addton to cultvar effects on coffee converson ratos (Temesgen and Mchael, 1995), cultural practces to ncrease coffee bean weght n other coffee cultvars may enhance cherry to green coffee converson yelds. For example, mulchng (Kamau, 1976), optmal balance of nutrents (Njoroge, 1998) or shade (Muschler, 2000) may beneft coffee ndustral yeld conversons. In contrast, Bourbon and Yellow Caturra followng wth hgh absolute scores (Table 3) requred more cherry frut to completng 46 kg (1 Qq) of green coffee those dates (Table 1). As wth Typca, no clear advantages were observed for these cultvars wth respect to complete 46 kg (1 Qq) of green coffee from parchment (Table 1). The CDF 2 accounted for only 8.0% of the varaton among the coffee cultvars; therefore, ths functon was not used for nterpretaton. A plot of the canoncal coeffcent scores (Fgure 2) of the coffee cultvars llustrates the hgh dscrmnant capacty of CDFj n separatng Typca and Bourbon from the other cultvars, manly on the bass of ther cherry to green coffee conversons n January, February, March 1999, and May Thus on the bass of the lnear combnaton presented n Equaton 3, the complexty of the data was summarzed. Ths model was useful to understand coffee converson yelds on several harvest dates carred out wth eght dfferent coffee cultvars. CDFj = 1.29x x x x x x x x x x x n x 12 ( " 3 ' ) Where x 1; x 2,..., x 12 represent the cherry and parchment to green coffee converson factors descrbed n Table 2. In general, Bourbon and Yellow Caturra yeld more frut than Typca and Mundo Novo (Vllaseñor, 1987). However, n the present work Typca showed hgher cherry to green coffee conversons because the sze of the beans was larger. Therefore, local cherry producer growers would receve premum prces from the coffee processor ndustry for Typca and Mundo Novo frut because of the hgh converson rate of cherry to green coffee, and less reward for cherres harvested from

10 178 ESPINSA-SLARES ET AL./CFFEE PRCESSING YIELD u q oo Q U J» A Catua D Pluma Hdalgo G a mca Typca Red Caturra Yellow Caturra Bourbon Mundo novo A A A t « CDF, (83.2 %) FIGURE 2. Plot of the canoncal scores of the frst two canoncal dscrmnant functons for cherry to green, and parchment to green coffee conversons harvested on dfferent dates n 1999 and 2000 for eght dfferent coffee cultvars. Bourbon and Yellow Caturra. The decson on plantng cultvars wll depend on the grower market strategy commercalzng cherry, parchment or green coffee. LITERATURE CITED Alpzar, J. M., Comportamento de los rendmentos de benefcado de café de Costa Rca por zonas de produccón, según regstros de 15 años. pp In: L. Zamora and J. Echeverr (ed.), XIX Smposo Latnoamercano de Cafcultura, Memora, ICAFE, San José, Costa Rca. Atmawnata,., The safe storage mosture content for green coffee. Pelta Perkebunan 11: Castllo, C. F, Evaluacón de Catmor Brote Bronce y Caturra Crollo desde el crecmento hasta la productvdad, pp In: Smposum Latnoamercano sobre Cafcultura, Guatemala. Cléves-Serrano, R., Factores que ncden en la caldad del café, orgnados en el cultvo y la recoleccón, pp In: R. Cléves-Serrano (ed.), Tecnología en benefcado de café, San José, Costa Rca. Cruz-Castllo, J. G., S. Ganeshanandam, B. R. MacKay G. S. Lawes, C. R.. Lawoko and D. J. Woolley Applcatons of canoncal dscrmnant analyss n hortcultural research. HortScence 29(10): Fgueroa-Solares, P.,. H. Jménez, E. López de León and F Anzueto, Influenca de la varedad y la alttud en las característcas organoléptcas y físcas del café.

11 J. Agre. Unv. P.R. VL. 89, N. 3-4, JULY-CTBER pp In: L. Zamora and J. Echeverr (ed.), XIX Smposo Latnoamercano de Cafcultura, Memora, ICAFE, San José, Costa Rca. Gobb, J. A., Is bodversty-frendly fnancally vable? An analyss of fve dfferent coffee producton systems n western El Salvador. Ecologcal Economcs 33: González-Arcos, E, Determnacón de rendmentos en benefco de café Coffea arábca, en dferentes alttudes de la regón del Soconusco, Chapas, Tess Ingenero Agrónomo. Ftotecna-UACh, Chapngo, Méxco, pp Guadarrama, Z. C. and. L. Trujllo, Caldad de café, varedades y fertldad del suelo entre pequeños productores de Ixhuatlan del Café, Ver.: dagnóstco prelmnar. Memoras XIV Reunón Centífca-Tecnológca Forestal y Agropecuara Veracruz, Ver ctubre. Méxco. Guyot, B., D. Gueule, J. C. Manez, J. J. Perrot, J. Grón, and L. Vllan, Influence de l'alttude et de l'ombrage sur la qualte des cafes arábca. Plantatons Recherche Developpement 3: Haarer, A. E., Produccón moderna de café. 4- mpresón. Méxco. 652 pp. Hmmelblau, D. M., Prncpos Báscos y Cálculos en Ingenería Químca. Pentrce Hall Latnoamercana, Méxco. 728 pp. Kamau, I. N, Effects of cultural practces on the qualty of Kenya coffee Part I. Effects of Naper grass mulch and ntrogen fertlzers on the qualty of arábca coffee n Kenya. Kenya Coffee 41: Krzanowsk, W. J., Prncples of multvarate analyss. A user's perspectve. xford, Clarendon. 380 pp. Ky, L. Louarn, J. Dussert, S. Guyot, B. S. Hamon and M. Norot, Caffene, trgonellne, chologenc acds and sucrose dversty n wld Coffea arábca L. and C. canephora P. accessons. Food Chemestry 75: Martn, M. J., P. F González, A. G. Valdenebro and M. León-Camacho, Fatty acd profles as dscrmnant parameters for coffee varetes dfferentaton. Talanta 54: Montagnon, Ch., T Ch. Leroy, A. Yapo and P. Charmetant, Genotype-locaton nteractons for Coffea canephora yeld n the Ivory Coast. Agronome 20: Muschler, R. G., Shade mproves coffee qualty n a sub-optmal zone of Costa Rca. pp In: L. Zamora and J. Echeverr (ed.), XIX Smposo Latnoamercano de Cafetcultura, Memora, ICAFE, San José, Costa Rca. Njoroge, J. M., Agronomc and processng factors affectng coffee qualty. utlook on Agrculture 27: Puerta-Quntero, G. I., 2000a. Caldad en taza de algunas mezclas de varedades de café de la espece Coffea arábca L. Cencafé 51:5-19. Puerta-Quntero, G. I., 2000b. Influenca de los granos de café cosechados verdes en la caldad físca y organoléptca de la bebda. Cencafé 52: Reyna, F, Evaluacón de cnco varedades de café Coffea arábca Zona Río Arrba, El Paraíso, pp In: XII Smposo de Cafetcultura Latnoamercana, San Pedro Sula, Honduras. SAS Insttute, SAS/STAT User's Gude. Versón 6. 4th edton. Cary N Soto-Pnto, L., R I. Castllo-Hernández and J. Caballero-Neto, Shade effect on coffee producton at northern Tzeltal zone of the state of Chapas, Mexco. J. Agrculture Ecosystem Envronment 80: Steel, R. G. D. and J. H. Torre, Prncples and Procedures of Statstcs: A Bometrcal Approach. McGraw-Hll, New York. 633 pp. Temesgen, G. and G. W. Mchael, Determnaton of converson ratos of fresh cherry to clean coffee on promsng coffee cultvars at Gera. IAR Newsletter of Agrcultural Research 10:9-12.

12 180 ESPINSA-SLARES ET AL./CFFEE PRCESSING YIELD van der Stegen, G. H. D., Enhancement of coffee qualty by mould preventon. Food Control 14: Vlla, A., Cultvo del Café en Méxco, Edtoral La Fuente SA, Méxco, pp Vllaseñor, L. A., Cafetcultura Moderna en Méxco. Edtoral Futura, Méxco. 467 pp.