1 Dispersion Patterns of the American Leaf Spot ( Mycena citricolor) in Costa Rican Coffee Systems Rebecca Spicer Warner College of Natural Resources, Colorado State University ABSTRACT Coffee ( Coffea arabica ) is both economically and culturally important to Costa Rica; however it is also highly susceptible t o invasion by pests and pathogens. One such fungal p athogen is the American Leaf Spot ( Mycena citricolor) This fungus thrives in the same high-elevation, humid habitats that produce some of the highest quality coffee. P lants harboring the fungus in great loads typically experience defoliat ion and fruit loss. The dispersion patterns of M. citricolor were investigated in eight plots located in Caitas, Cos ta Rica. The number of fungal spots present was re corded for a total of 200 coffee plants, and the dispersion patterns were des cribed for each plot and for the area. When compar ed to the expected Poisson distribution, five plots exhibited a patchy distribution and three plots were random. Fungal dispersion across all plots was also patchy. The predominance of the patchy di stribution is related to the life cycle of the fung us, whereas management practices likely account for deviation from this pa ttern. RESUMEN El caf es importante en la economa y la cultura d e Costa Rica, sin embargo es muy sensible a la inva sin de plagas y enfermedades. Una de estas enfermedades es el hong o ojo de gallo ( Mycena citricolor ). Este hongo crece bien en las mismas zonas altas y ambientes hmedos donde se pr oduce caf de calidad buena. Las plantas que tiene n este hongo en cantidades grandes tpicamente muestran defoliacin y prdida de frutos. Los patrones de dispersin fueron investigados en ocho terrenos en Caitas, Costa Rica. El nmero de manchas por el hongo fue anotado para un total de 200 plantas del caf, y los patrones de dispersin fueron descritos pa ra cada terreno y para el rea. Cuando se compar la distribucin Poisson, cinco terrenos mostraban una distribucin agregada y tres terrenos eran aleatorios. La dispersin del hongo en todos los terrenos tambin era agregado. La predominancia de la distribucin agregada est relacionada al ciclo de vida del hongo, mientras diferentes prcticas de mantenimiento prob ablemente explican la desviacin de este patrn. INTRODUCTION Coffea arabica known commonly as the coffee plant, is integral t o both the past and present of Costa Rica (Boucher 1983). This traditional crop is now the third most important crop in an export economy worth over $9 billion in 2007 (CIA 2008). Although tourism has overtaken agriculture in terms of percent GDP generated (CIA 2008), coffee remains a significant contributor to the Costa Rican economy through popular Â“coffee tours.Â” The coffee industry has been further impacted by the food certification movement, and coffee producers nation wide are faced with the choice of growing traditional, organic, sustainable or fair trade cof fee (Griffith et al. 2000). Considering both the cultural and economic signifi cance of the coffee plant, much attention is given to maintaining a healthy crop. Fortunately, during its 200 years of cultivation in Central Amer ica there have been very few serious pest problems. Di seases such as leaf rust ( Hemileia vasatrix ) and the berry borer ( Hypothenemus hampeii ) are kept at minimal levels by the shaded environm ent of traditional coffee systems, and the coffee leaf miner ( Leucoptera coffeella ) is controlled by infrequent weed slashing. An exception to this phenomenon has been the case of the American Leaf Spot, Mycena citricolor (Basidiomycota). Ironically, the same conditions that have kept other fungal pathogens at bay and have been credited with producing high quality coffee have also shown to be ideal for the growth and propagation of M. citricolor Known throughout Central America as ojo de gallo this basidiomycete thrives in habitats characterized by high humidity and with more or less continuous rain s,
2 typically areas of upper elevation (Staver et al. 2 001). Studies have found that the fungus is most widespread in regions between 1,100 and 1,500 m (Av elino et al. 2007), and laboratory inoculations proved most successful at 20C (Rao and Tewari 1987 ). Heavy clouds and fog in these regions limit light penetration and air circulation, processes th at ordinarily would allow leaf surfaces to dry (Sta ver et al. 2001). Mycena citricolor infests coffee plants in the form of a compact mass of hyphae (filaments). The head of each hypha is a propagule, or gemma. Gemma e become detached with contact (often human) or in the presence of water. Germination of the gemma e constitutes asexual reproduction for the fungus (Alexopoulos 1996, Wellman 1950, in Avelino et al. 2007). The fungus can attack leaves, branches or berries. Infected leaves are characterized by ligh t-brown circular spots between five and ten mm in diameter. Defoliation follows infection, in additi on to fruit fall in more severe cases (Staver et al 2001). One explanation for the severity of the damage done in Central American plantations is that this hostpathogen system has not coevolved. Rather, M citricolor is of American origin and coffee is African (Ploetz 2007). A recent study showed that M. citricolor can be best controlled by maintaining light shadin g, spacing rows farther apart than what is found in tr aditional systems and pruning the coffee plants (Avelino et al. 2007). Despite the simple and inex pensive nature of these recommendations, management techniques for this fungus can often con flict with methods used to control other common fungi. This discrepancy typically derives from the conditions created by shade-grown coffee systems. Microclimatic conditions beneath shade trees are ty pically cooler and more humid than more modern sun-grown systems (Barradas and Fanjul 1986). Hemileia vasatrix and M. citricolor thrive beneath shade (Avelino et al. 2004), but Cercospora coffeicola actually suffers under these conditions (Staver et al. 2001). Thus farmers must choose a technique ba sed on which fungus type they wish to manage. In many cases fungicides are also used to control what they otherwise cannot (Avelino et al. 2007). When investigating the presence of a fungal pathog en such as M. citricolor one parameter that is commonly studied is the dispersion pattern. Disper sion patterns refer to the probability that, given the location of one individual, another individual is n earby. They typically fall into one of three cate gories: patchy, uniform or random. A patchy distribution i s characterized by a high probability of another individual being nearby, whereas a uniform distribu tion would have a low probability. The probability is unaffected in a random distribution (Krebs 1999) The purpose of this study was to examine dispersio n patterns of M. citricolor in coffee systems located in the Monteverde area of Costa Rica. The elevation, moisture and temperature conditions of this area make it a prime location for the study of the fungus. If the dispersion of M. citricolor can be described, management of agricultural systems can b e designed to reduce damage caused by its infestation. Given that reproduction of the fungus occurs on a localized scale, I predict that disper sion will be patchy. METHODS Study Sites This study was performed from 22 April 7 May, 200 8 on five coffee farms in Caitas, Puntarenas, Costa Rica (Table 1). Mycena citricolor was present on all farms, although methods of contr ol varied. Atemi (cyproconazole) was the most common fungicide used though some landowners were seeking more organic methods of fungal control. All coffee farms were planted alongside other productive crops, often sugar cane, avocado, banana and citrus From these five farms, eight coffee plots were defined, ranging from 0.25 to two hectares in size. Each plot was at least 50 m away from the nearest plot and was separated by non-coffee habitat.
3 Sampling Procedures At each study plot 25 coffee plants were sampled. To ensure that the sample was representative, plant s were arbitrarily chosen from locations throughout t he entire plot, and only plants between one and 1.5 m tall were sampled. For each plant the number of le aves harboring the fungus and the total number of fungal spots were recorded. Note that a coffee Â“pl antÂ” is considered what has been planted in a singl e hole and traditionally consisted of two distinct in dividuals. Statistical Analysis Frequency distributions of each individual plot and of the total sample were compared to a Poisson distribution in order to determine if spatial patte rns observed were random. A chi-squared goodness-o ffit test was used to compare the expected random di stributions to the observed distributions. In addi tion, the fungal counts for all plots were also summed an d the same tests were run for the area as a whole. The index of dispersion was also calculated for eac h plot as the variance divided by the average numbe r of fungal spots per plant. These values were used to determine to what degree plots could be consider ed patchy or uniform. RESULTS Fungal distribution Dispersion patterns varied from random to nonrandom between the eight sample plots, but M. citricolor was dispersed nonrandomly when all data points were compiled into a single frequency distribution. Three plots did not vary significantly from the pre dicted Poisson: DJ (chi-squared = 7.09, df = 5, p > 0.05), ST3 (chi-squared = 1.19, df = 5, p > 0.05) a nd MB (chi-squared = 5.07, df = 3, p > 0.05). The remaining five plotsST1 (chi-squared = 36.41, df = 5, p < 0.05), ST2 (chi-squared = 8.11, df = 3, p < 0.05), SM (chi-squared = 5.58, df = 1, p < 0.05), T T1 (chi-squared = 8.62, df = 4, p < 0.05) and TT2 (chi-squared = 17.44, df = 3, p < 0.05)varied sig nificantly from the expected Poisson distribution a nd exhibited a nonrandom distribution (Figures 1-8). Scaling up to the regional dispersion pattern, the total fungal counts for all plots also yielded a nonrando m distribution (Figure 9, chi-squared = 274.85, df = 7, p < 0.05). Indices of dispersion calculated were above one fo r each plot, indicating that not only were the observed distributions nonrandom, but they were als o patchy. Furthermore indices varied between each plot revealing that this patchiness was not equally patchy. (Figure 10). At 170.30, Plot ST1 had the highest index of dispersion, indicating that the ra nge of values for spots per plant was many times greater than the average number of spots per plant. Conversely, the variance and mean for SM are more similar and the plot has a lower index of dispersio n (7.23). All other values fell between these two. Additional observations I noted that in most cases the fungus was more comm on on the lower branches of the plants than on middle or upper branches, though there were excepti ons. Also I found that the most severely infested trees were found on the edges of the plots, most ne arest to other plant species and most likely to be shaded. Plants with low infestation rates were com monly found in full sun, often in the center of the rows. All plots planted on slopes (ST1, ST3, TT1) were south-facing, maximizing sunlight on these plots. All farms practiced intercropping with the most common crops being sugar cane, banana, citrus trees and avocado. In addition all plots were plan ted with windbreaks, and non-native plants were use d on ST3 (cypress), MB (cypress and pine) and TT1 (eu calyptus).
4 DISCUSSION My prediction that the dispersion of M. citricolor would be patchy was supported in five of eight plots sampled. The simplest explanation for the patchy d ispersion comes from the dispersal mechanism of the fungus. Gemmae are only dispersed through direct c ontact and the flow of water, thus it follows that the fungus should spread most easily within a plant or to nearby plants. In a natural system, fungi dispe rsed by this mechanism would likely exhibit a patchy dis persion, and this was true in most plots. Three plots, however, showed a random dispersion. The farms where random patterns were found may have had management or agricultural pract ices that affected the dispersion of the fungus. They were owned by different people and differed in many variables, including principle use (Table 1). For instance, Plot DJ is used principally as a tour ist attraction and uses a wide variety of fungicide s. Efforts to keep it more Â“presentableÂ” may have elim inated the patchy nature of the fungus. The factor s influencing the other random plotsST3 and MBare less clear. The only distinguishing factor of the se plots was the use of non-native gymnosperms as wind breaks. Though the mechanism is not understood, these plants may have disrupted typical dispersion patterns. Though not tested here, it is possible that the ag e of the coffee plants in a plot could affect the dispersion patterns of M. citricolor In young plots, few plants may be infected and t herefore the dispersion may appear random. As time proceeds and the fungus spreads, the patchy nature of its dispersion may become evident. Other factors that may affect M. citricolor dispersion include the microenvironment, crop management and characteristi cs of the coffee trees themselves. As discussed previously, infestation of M. citricolor is tightly correlated with the abiotic conditions p resent. Avelino et al. (2007) found that multiple factors were sign ificant, resulting in a highly complex pathosystem. They specifically emphasized the significance of to pography, especially slope exposure and inclination and the detrimental impacts of shade, specifically that provided by fruit and forest trees. Both of t hese could have been critical to the Caitas study, but again, the data do not allow this to be concluded. The degree of patchiness was also shown to vary be tween the eight plots. These results are more easily explained by conditions within the plots. F or example, plot ST1 had the highest index of dispersion, that is the highest degree of dispersio n. This plot was located on a steep, S-facing slop e surrounded by tall windbreak trees. The high degre e of variance is largely due to the contrasting lev els of sun and shade and the resulting microclimate con ditions; the centers of coffee rows were in direct sunlight for most of the day, but edge plants where almost always shaded. This caused large differenc es in fungal counts per plant. On the other end, plot SM had very low counts as it was the smallest plot sampled and was primarily homogenous. It follows t hen that plot heterogeneity may magnify the patchiness displayed by M. citricolor Nonetheless, on a large, multi-plot scale the data showed that M. citricolor typically does exhibit a patchy distribution. At this scale it would appe ar that local effects such as crop and windbreak va riety and fungicide use were overshadowed by the natural behavior of this fungus Another opposing but intriguing theory is that M. citricolor does not display a typical or natural dispersion p attern. Rather the dispersion pattern is dynamic through time, constan tly adjusting to year-to-year climatic conditions. Considering that most farmers indicated the use of fungicides only in particularly rainy years, it is evident that attributes of the M. citricolor pathosystem do vary temporally, and it is likely t hat the dispersion pattern may be one of these attributes. Though the results of this study were straightforwa rd, there are many aspects of fungal dispersion in coffee systems that are worthy of fur ther study. In order to understand better the behav ior of M. citricolor similar procedures should be completed throughout the range of the fungus in order to determine if the observed patchiness was a regional phenomenon or if it occurs in all environments. T he temporal variation in dispersionin one plot over time or in different aged plotswould be a worthwh ile exploration of dispersion pattern development or th e dynamics of dispersion patterns through time. It
5 would also be interesting to examine fungal dispers ion in relation to fungal severity to see if the tw o factors are correlated. Considering that multiple fungal pathogens are found in coffee plantations, further studies could also compare the dispersion p atterns between species. ACKNOWLEDGEMENTS First and foremost, I would like to thank Karen Mas ters for advising me (a lot) throughout this projec t, especially when my initial idea did not work out. Additional thanks t o Alan Masters for sparking my interest in a study on coffee and Pablo Allen for his assistance with the Poisson distribution. I am very thankful for the revisions provided by Ka therine Heal and for the translations assistance from Moncho Caldern. Most importantly, this project would not have been poss ible without my study sites. I am indebted to Luis Alonso Cruz, Vi ctor Torres, Edwin Santamara, Hannia Zamora Mirand a and Juan de Dios for the generous use of their beautiful farms. LITERATURE CITED Alexopoulos, C.J., C.W. Mims, and M. Blackwell. 19 96. Introductory Mycology John Wiley & Sons, Inc., New York, NY, pp. 27, 700. Avelino, J., S. Cabut, B. Barboza, M. Barquero, R. Alfaro, C. Esquivel, J. Durand, and C. Cilas. 2007 Topography and crop management are key factors for the development of American Leaf Spot epidemics on coffee in Costa Rica. Phytopathology 97: 1532-1542. Avelino, J., L. Willocquet, and S. Savary. 2004. Effects of crop management patterns on coffee rust epidemics. Plant Pathology 53: 541-547. Barradas, V.L. and L. Fanjul. 1986. Microclimatic characterizations of shaded and open-grown coffee ( Coffea arabica ) plantations in Mexico. Agricultural and Forestry Meteorology 38: 101-112. Boucher, D.H. 1983. Coffee. In: Costa Rican Natu ral History D.H. Janzen, ed. The University of Chicago Press, Chicago, IL, pp. 86-88. Central Intelligence Agency. 2008. The World Fact book: Costa Rica. https://www.cia.gov/library/publications/the-world -factbook/geos/cs.html. Accessed 13 March 2008. Griffith, K., D.C. Peck, and J. Stuckey. 2000. Ag riculture in Monteverde: moving towards sustainability. In Monteverde: Ecology and Conser vation of a Tropical Cloud Forest N.M. Nadkarni and N.T. Wheelwright, eds. Oxford Univer sity Press, New York, NY, pp. 389-408. Krebs, Charles J. 1999. Ecological Methodology Addison-Wesley Educational Publishers, Inc., pp. 115-123. Ploetz, R.C. 2007. Diseases of tropical perennial crops: challenging problems in diverse environment s. Plant Disease 91:644-663. Rao, D. V., and J.P. Tewari, J. P. 1987. Production of oxalic acid by Mycena citricolor causal agent of the American leaf spot of coffee. Phytopathology 77:780-785. Staver, C., F. Guharay, D. Monterroso, and R.G. Mus chler. 2001. Designing pest suppressive multistrata perennial crop systems: shade-grown co ffee in Central America. Agroforestry Systems 53: 151-170. Wellman, F. L. 1950. Dissemination of Omphalia leaf spot of coffee. Turrialba 1:12-27.
6 TABLE 1. Characteristics of the farms used as study sites. Farms were all located in Caitas, Costa Rica, but differed in age, size, distances between plots, uses and fungal treatments. Finca Don Juan Sociedad El Tabacn Finca Santamara Finca Marbella Tour de Trapiche Plots DJ ST1, ST2, ST3 SM MB TT1, TT2 Age (in years) 3 18 15 15 >30 Area per plot (in hectares) 2 1.25, 0.75, 0.5 0.25 2 4, 0.25 Distance between plots (in meters) 300 50 200 200 100 Principle Use Tours Production Production Production TT1Product ion TT2Tours Fungicide use Atemi, boron, zinc Atemi Atemi None Lime Pruning yes yes yes yes yes
7 0 1 2 3 4 5 01-2526-5051-7576100 101125 126150 151175 176200 201225 226250 251275 276300 301325 326350 351375 376400 401425 Number of fungal spots per plantNumber of plants FIGURE 1. Number of M. citricolor spots on the leaves of 25 coffee plants found on t he Don Juan coffee farm in Caitas, Costa Rica. 0 1 2 3 4 5 6 7 8 9 10 01-2526-5051-7576100 101125 126150 151175 176200 201225 226250 251275 276300 301325 326350 351375 376400 401425 Number of fungal spots per plantNumber of plants FIGURE 2. Number of M. citricolor spots on the leaves of 25 coffee plants found on t he first of three plots (ST1) in Sociedad el Tabacn in Caitas, Cost a Rica. Average # of spots per plant = 108.84 Average # of spots per plant = 76.12
8 0 2 4 6 8 10 12 01-2526-5051-7576100 101125 126150 151175 176200 201225 226250 251275 276300 301325 326350 351375 376400 401425 Number of fungal spots per plantNumber of plants FIGURE 3. Number of M. citricolor spots on the leaves of 25 coffee plants found on t he second of three plots (ST2) in Sociedad el Tabacn in Caitas, Cost a Rica. 0 1 2 3 4 5 01-2526-5051-7576100 101125 126150 151175 176200 201225 226250 251275 276300 301325 326350 351375 376400 401425 Number of fungal spots per plantNumber of plants FIGURE 4. Number of M. citricolor spots on the leaves of 25 coffee plants found on t he third of three plots (ST3) in Sociedad el Tabacn in Caitas, Cost a Rica. Average # of spots per plant = 50.60 Average # of spots per plant = 154.44
9 0 5 10 15 20 25 01-2526-5051-7576100 101125 126150 151175 176200 201225 226250 251275 276300 301325 326350 351375 376400 401425 Number of fungal spots per plantNumber of plants FIGURE 5. Number of M. citricolor spots on the leaves of 25 coffee plants found on F inca Santamara in Caitas, Costa Rica. 0 1 2 3 4 5 6 7 8 9 01-2526-5051-7576100 101125 126150 151175 176200 201225 226250 251275 276300 301325 326350 351375 376400 401425 Number of fungal spots per plantNumber of plants FIGURE 6. Number of M. citricolor spots on the leaves of 25 coffee plants found on F inca Marbella in Caitas, Costa Rica. Average # of spots per plant = 9.64 Average # of spots per plant = 47.16
10 0 1 2 3 4 5 6 7 8 01-2526-5051-7576100 101125 126150 151175 176200 201225 226250 251275 276300 301325 326350 351375 376400 401425 Number of fungal spots per plantNumber of plants FIGURE 7. Number of M. citricolor spots on the leaves of 25 coffee plants found on t he first of two plots (TT1) at the Tour de Trapiche coffee farm in Caitas, Costa Rica. 0 2 4 6 8 10 12 14 16 18 20 01-2526-5051-7576100 101125 126150 151175 176200 201225 226250 251275 276300 301325 326350 351375 376400 401425 Number of fungal spots per plantNumber of plants FIGURE 8. Number of M. citricolor spots on the leaves of 25 coffee plants found on t he second of two plots (TT2) at the Tour de Trapiche coffee farm in Caitas, Costa Rica. Average # of spots per plant = 70.96 Average # of spots per plant = 17.56
11 0 10 20 30 40 50 60 70 01-2526-5051-7576100 101125 126150 151175 176200 201225 226250 251275 276300 301325 326350 351375 376400 401425 Number of fungal spots per plantNumber of plants FIGURE 9. Number of M. citricolor spots on the leaves of coffee plants sampled on a total of eight plots in Caitas, Costa Rica (N = 200). FIGURE 10. Indices of dispersion for eight coffee plots are calculated as the variance over the avera ge number of M. citricolor spots found per plant. Larger values for the index indicate that variance was many times greater than the mean. Plot heterogenei ty generally led to a higher index of dispersion. 0 20 40 60 80 100 120 140 160 180 SMST3TT2TT1DJST2MBST1 Coffee plotIndex of dispersion Average # of spots per plant = 66.92
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xlink http:www.w3.org1999xlink xmlns:xsi http:www.w3.org2001XMLSchema-instance
leader 00000nas 2200000Ka 4500
controlfield tag 008 000000c19749999pautr p s 0 0eng d
datafield ind1 8 ind2 024
subfield code a M39-00023
Dispersin de los patrones del hongo ojo de gallo (Mycena citricolor) en los sistemas de caf de Costa Rica
Dispersion patterns of the American Leaf Spot (Mycena citricolor) in Costa Rican coffee systems
Coffee (Coffea arabica) is both economically and culturally important to Costa Rica; however it is also highly susceptible to invasion by pests and pathogens. One such fungal pathogen is the American Leaf Spot (Mycena citricolor). This fungus thrives in the same high-elevation, humid habitats that produce some of the highest quality coffee. Plants harboring the fungus in great loads typically experience defoliation and fruit loss. The dispersion patterns of M. citricolor were investigated in eight plots located in Caitas, Costa Rica. The number of fungal spots present was recorded for a total of 200 coffee plants, and the dispersion patterns were described for each plot and for the area. When compared to the expected Poisson distribution, five plots exhibited a patchy distribution and three plots were random. Fungal dispersion across all plots was also patchy. The predominance of the patchy distribution is related to the life cycle of the fungus, whereas management practices likely account for deviation from this pattern.
El caf es importante en la economa y la cultura de Costa Rica, sin embargo es muy sensible a la invasin de plagas y enfermedades. Una de estas enfermedades es el hongo ojo de gallo (Mycena citricolor). Este hongo crece en las mismas alturas elevadas y ambientes hmedos donde se produce el caf de buena calidad. Las plantas que tienen este hongo en cantidades grandes tpicamente muestran defoliacin y prdida de frutos. Los patrones de dispersin fueron investigados en ocho terrenos en Caitas, Costa Rica.
Text in English.
Coffee--Diseases and pests--Costa Rica--Guanacaste--Caitas
Caf-Enfermedades y plagas--Costa Rica--Guanacaste--Caitas
Tropical Ecology 2008
American leaf spot disease of coffee
Ecologa Tropical 2008
Enfermedad Ojo de gallo
t Monteverde Institute : Tropical Ecology