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 M39-00350
Regeneracin de los bosques en las zonas infectadas de Rosellinia de Los Llanos, Monteverde y los agentes de control biolgicos posibles
Forest regeneration in Rosellinia infected areas of Los Llanos, Monteverde, and possible biocontrol agents
A fast-spreading fungus in the genus Rosellinia threatens the forest composing the Monteverde Corridor in Los Llanos area. This fungus spreads through the soil and has been observed to kill both adult and juvenile trees. The intent of this study was to learn more about the effects of Rosellinia in the forest with the ultimate goal of creating a biocontol. The experiments consisted of four parts: seed germination of pioneer and canopy tree species, sapling infection rates of pioneer and canopy tree species, evaluation of the understory composition of the infected area in comparison to a pristine, and previous tree composition data, and the investigation of the presence of another fungus from the genus Trichoderma that may serve as a natural predator of Rosellinia. Overall, the fungus was found to be a general root specialist. Of the two genera evaluated as possible biocontrols, Bocconia was found to have effects on the germination of Ocotea
whitei seeds in infected soil; however, these effects did not extend to Citharexylum costaricensis seeds. The absence of Trichoderma from the soil does not seem to be the reason Rosellinia established, nor does it look promising as a biological control without the introduction of an exotic species. Evaluation of the regenerating plant community in the infected area indicated that Rosellinia has not yet affected the seed
bank. This fact gives hope for the future in that if a biological control can be found, the forest should be able to regenerate to a pristine condition.
Un hongo en el gnero Rosellinia, que se extiende muy rpidamente, amenaza al bosque que compone el Pasillo de Monteverde en el rea de Los Llanos. Este hongo se esparce por el suelo y se ha observado que mata a los rboles adultos y juveniles. El objetivo de este estudio fue aprender ms acerca de los efectos de Rosellinia en el bosque, con el fin de crear un biocontol. Los experimentos consistieron de cuatro partes: la germinacin de las semillas de especies de rboles pioneros y de dosel, las tasas de infeccin de plntulas de especies de rboles pioneros y de dosel, la evaluacin de la composicin del sotobosque del rea infectada con respecto a los datos de composicin de rboles prstinos y previos de la composicin del rbol, y la investigacin de la presencia de otro hongo del gnero Trichoderma que puede servir como un depredador natural de Rosellinia. En trminos generales, se encontr que el hongo fue un especialista general de races. Se determin que Bocconia, entre los dos gneros evaluados como posibles agentes de biocontrol, afect la germinacin de semillas de Ocotea whitei en los suelos infectados; sin embargo, estos efectos no se extendieron a las semillas de Citharexylum costaricensis. La ausencia de Trichoderma en el suelo no parece ser la razn por la que Rosellinia se estableci, ni tampoco muestra posibilidades como un control biolgico sin la introduccin de una especie extica. La evaluacin de la comunidad vegetal en regeneracin en el rea infectada indic que Rosellinia an no haba afectado el banco de semillas. Este hecho da esperanzas para el futuro ya que si un control biolgico puede ser encontrado, entonces el bosque debe ser capaz de volver a una condicin prstina.
Text in English.
Costa Rica--Puntarenas--Monteverde Zone--Los Llanos
Regeneracin del bosque
Costa Rica--Puntarenas--Zona de Monteverde--Los Llanos
Tropical Ecology Spring 2005
Ecologa Tropical Primavera 2005
t Monteverde Institute : Tropical Ecology
1 Forest regeneration in Rosellinia infected areas of Los Llanos, Monteverde, and possible biocontrol agents Jessica Smith Department of Biology, University of Wisconsin Madison ABSTRACT A fast spreading fungus in the genus Rosellinia threatens the for est composing the Monteverde Corridor in Los Llanos area. This fungus spreads through the soil and has been observed to kill both adult and juvenile trees. The intent of this study was to learn more about the effects of Rosellinia in the forest with the ultimate goal of creating a biocontol. The experiments consisted of four parts: seed germination of pioneer and canopy tree species, sapling infection rates of pioneer and canopy tree species, evaluation of the understory composition of the infected area in comparison to a pristine, and previous tree composition data, and the investigation of the presence of another fungus from the genus Trichoderma that may serve as a natural predator of Rosellinia . Overall, the fungus was found to be a general root spec ialist. Of the two genera evaluated as possible biocontrols, Bocconia was found to have effects on the germination of Ocotea whitei seeds in infected soil; however, these effects did not extend to Citharexylum costaricensis seeds. The absence of Trichoder ma from the soil does not seem to be the reason Rosellinia established, nor does it look promising as a biological control without the introduction of an exotic species. Evaluation of the regenerating plant community in the infected area indicated that Ros ellinia has not yet affected the seed bank. This fact gives hope for the future in that if a biological control can be found, the forest should be able to regenerate to a pristine condition. RESUMEN Un hongo en el gÃ©nero Rosellinia, que se extiende muy rÃ¡pidamente, amenaza al bosque que compone el Pasillo de Monteverde en el Ã¡rea de Los Llanos. Este hongo se esparce por el suelo y se ha observado que mata a los Ã¡rboles adultos y juveniles. El objetivo de este estudio fue aprender mÃ¡s acerca de los efecto s de Rosellinia en el bosque, con el fin Ãºltimo de crear un biocontol. Los experimentos consistieron de cuatro partes: la germinaciÃ³n de las semillas de especies de Ã¡rboles pioneros y de dosel, las tasas de infecciÃ³n de plÃ¡ntulas de especies de Ã¡rboles pio neros y de dosel, la evaluaciÃ³n de la composiciÃ³n del sotobosque del Ã¡rea infectada con respecto a datos de composiciÃ³n de Ã¡rboles prÃstinos y previos de la composiciÃ³n del Ã¡rbol, y la investigaciÃ³n de la presencia de otro hongo del gÃ©nero Trichoderma que puede servir como un depredador natural de Rosellinia. En tÃ©rminos generales, se encontrÃ³ que el hongo fue un especialista general de raÃces. Se determinÃ³ que Bocconia, entre los dos genÃ©ros evaluados como posibles agentes de biocontrol, afectÃ³ la germina ciÃ³n de semillas de Ocotea whitei en suelos infectados; sin embargo, estos efectos no se extendieron a las semillas de Citharexylum costaricensis . La ausencia de Trichoderma en el suelo no parece ser la razÃ³n por la que Rosellinia se estableciÃ³, ni tampoco muestra posibilidades como un control biolÃ³gico sin la introducciÃ³n de una especie exÃ³tica. La evaluaciÃ³n de la comunidad vegetal en regeneraciÃ³n en el Ã¡rea infectada indicÃ³ que Rosellinia aÃºn no habÃa afectado el banco de semillas. Este hecho da esperanz as para el futuro ya que si un control biolÃ³gico puede ser encontrado, entonces el bosque debe ser capaz de volver a una condiciÃ³n prÃstina.
2 INTRODUCTION The forest of Los Llanos, Monteverde acts as an ecological corridor for altitudinal migratory birds such as the endangered Three Wattled Bell Bird (Stinson 2004). Consequently, the presence of a pathogenic fungus, identified in the Division Ascomycota, Family Xylariaceae, Genus Rosellinia , in this region is very concerning . The site of greatest infe ction, located on the property of Leonel Arguedas and Franklin Arce, is about 53 m 2 ; though evidence of spread outside this hotspot is noticeable. The infected area is characterized by several large, dead trees that have fallen or lost their canopy tops a nd several dead saplings with dried out leaves. The origin of this hotspot of Rosellinia in Los Llanos is unknown. The suspected initial vector of Rosellinia is an infected coffee plant from an outside area (Stinson 2004). Rosellinia has been documented in tropical forests (Alexopolos et. al . 1996), but it is much more common in monocultures (Anonymous 2003). The diversity of natural fungi predators located in tropical soil communities should make tropical forests more resistant to pathogenic fungi such as Rosellinia ; however, this does not seem to be the case in Los Llanos. Determining why and how Rosellinia has established so well in Los Llanos is key to developing control methods. The general mechanism accepted for the spread of Rosellinia is that it travels through dirt and fine root materials to infect the roots of the next host plant (Petrini 1993). A 2004 study by EAP student Sarah Stinson indicated that the fungus could spread up to three and a half meters in 30 days. This study also provided e vidence that horses may serve as vectors for the transmission of the fungus by distributing soil and root material in their hooves (Stinson 2004). This is very concerning because a horseback riding trail used several times daily by two local tourist busin esses passes directly through the infected site. Another reason for concern is that all traditional control practices in the forest and in agriculture have been ineffective in containing the spread of Rosellinia . In July of 2004, Leonel Arguedas dug a c ontrol ditch around the infected area to try and deter the spread of the pathogen; however, the effects of Rosellinia are visible outside of this area. The World Avocado Congress of 1995 presented evidence that solarization of infected avocado roots signi ficantly reduced the presence of Rosellinia necatrix from the upper layers of soil; however, the fungus remained viable in deeper areas, and re infection was expected to occur (Lopez and Herrera 1996). Other control methods such as soil drainage and limit ation of irrigation practices are also ineffective (Anonymous 2003). The first step to finding a control method for Rosellinia is to understand more about the natural history of the pathogen and how it affects the natural community. Observations have l ed to the belief that Rosellinia does not discriminate between host plant size, age, or species, but this hypothesis has not yet been tested. This led to two proposed models for Rosellinia : first, that it does not discriminate and affects all species of a ll sizes and ages, and second that it prefers large, canopy trees. The effects of Rosellinia on the regenerating community within the infected site are also unknown. Is the regenerating community of the infected site comparable to an uninfected, regenera ting area, or are less woody species re colonizing? Two different genera were investigated as possible control methods. Seeds in the genus Bocconia have shown anti fungal properties against several fungi species in the division Basidiomycota (Karen Maste rs, pers. comm.). The presence of a mature
3 Bocconia tree and several seedlings within the area of greatest infection led to the hypothesis that Bocconia could resist Rosellinia and has possibilities as a biocontrol. A second control investigated was a di fferent fungus in the Division Ascomycota, Family Hypocreales, genus Trichoderma . Trichoderma parasitizes other fungi as a food source; currently, researchers have been able to find a species of Trichoderma capable of controlling every pathogenic fungus f or which a biocontrol has been sought (Harman 2004). Besides its possibilities as a biological control, the natural or induced absence of Trichoderma from the soil in the infected area could explain the dominance of Rosellinia . This study investigated several questions regarding Rosellinia . Does the fungus discriminate between host plants? How does the fungus affect the regenerating community? Does Bocconia or Trichoderma have possibilities as control methods? And is the absence of Trichoderma the c ause the establishment of Rosellinia ? Based on observations regarding these questions, it was hypothesized that Rosellinia does not discriminated between host plants, that the woody regenerating community would be reduced in the infected area, that Tricho derma would not be found in the infected area, and that Bocconia would increase seed germination in infected soil. METHODS Study Organism The mycelium of the fungus in advanced stages is very evident on the roots of dead plants. As seen in Figure 1, t he mycelium is gray in color and has a wooly appearance. Infected roots are characterized by the lack of filamentous root material and the formation of dirt in small balls around the root structures. Study Site The majority of the research took plac e in a fragment of secondary forest in the Los Llanos area of Monteverde, Costa Rica, altitude 1250 m. Prior to this study, the approximately 53 m 2 area of infection was identified by INTA to contain Rosellinia . The infected area crosses over the proper ty line between the land of Leonel Arguedas and Franklin Arce. A series of prevention ditches surround the area from previous attempts to halt the spread of the fungus. Trees outside of the prevention ditch also show the characteristic signs of death by R osellinia dried out leaves, death of the tree, the top of the tree broken off, or the whole tree fallen. In one section of the infected area there is a large gap due to the death of several large trees, along with several dried out saplings about 2 m in he ight. A horseback riding trail passes directly through the infected area and still remains in use by two companies. A pristine patch of secondary growth located approximately 1000 m from the infected site was used as a site of comparison. Rosellinia ha s not been observed in this area. The next forest patch approximately 250 m up the horseback riding trail was also used as a site for comparison. This site, located on the property of Franklin Arce, is very densely forested but, more dead plants are prese nt than would be expected; the presence of Rosellinia has not yet been confirmed in this area.
4 Seed Germination Experiments To investigate the effect of Rosellinia on seed germination, plots using three types of treatments and two different species were created. Citharexylum costaricensis (Verbenaceae) and Ocotea whitei (Lauraceae) were the species examined. Citharexylum costaricensis species are medium sized, fast growing trees common to cloud forests (Haber, et al. 2000). Ocotea whitei are large ca nopy trees commonly found in Monteverde that are essential to the diet of the Resplendent Quetzal following the nesting season (Haber, et al. 2000). These species were utilized because they were both producing fruit at the time of the study and because th ey posed a good comparison between pioneer and canopy species. The first plot was a control containing soil and leaf litter from an area not affected by Rosellinia . The second plot contained soil from the infected area and a root covered in the mycelium of the fungus. The third plot consisted of the same material as the second plot, but with the addition of Bocconia seeds. An entire seed, ground up seeds, and an extract from the seeds were all added to the soil. For all treatments, plastic lined baskets were utilized and 100 of each type of seed was planted. All seed plots were kept in flat shady areas and watered daily. After three weeks all seeds were examined for signs of germination numbers were recorded. A 2 x 3 Contingency Table was used to analy ze the results. Seedling experiments Twelve seedlings of C. costaricensis and 12 seedlings of (Lauraceae) were obtained from the Bell Bird Institute and placed into the infected area. is a canopy species of the fa mily Lauraceae endemic to the Los Llanos area. The seedlings were planted in dirt from a non infected area contained by a plastic covering. Root material covered in the gray mycelium of Rosellinia were taken from dead trees from inside the infected area and placed in the soil with the saplings. The C. costaricensis seedlings were kept in partial sun and seedlings were kept in the shade; all were watered daily. After three weeks the roots of all seedlings were investigated for the presen ce of Rosellinia growth with the guidance of Leonel Arguedas. A 2 x 2 Contingency Table Analysis was used to evaluate these data. The search for Trichoderma To explore the diversity of Trichoderma populations in the soil, samples were taken from four di stinct sites the infected area, a part of the pristine forest patch about along the horseback riding trail within the infected area. The infected area and the horsebac k riding trail were assumed to contain Rosellinia in the soil based on the previous positive results for Rosellinia from INTA; the number and kind of species present still remains unknown. The pristine area was assumed to be Rosellinia free, and the suspe cted area was assumed to be an area threatened by Rosellinia . For all sites, the leaf litter was pushed aside, and approximately 250g of soil was colleted in a plastic
5 Ziploc bag. Eight total samples were taken in each area, with the exception of the tra il where 4 total samples were taken. Of these eight, 4 were taken as dry soil only, the other 4 had water and damp paper added to the bag. The samples were submitted to INTA lab in San Jose to identify the presence of Trichoderma . Understory Plant Com position To investigate the effects of Rosellinia on the regenerating plant community, samples from understory plants were collected exhaustively from the infected area and a disturbed area of similar size in the pristine area. Botanists Bill and Willow Haber helped to identify the family, genus, and species of the plants collected. The habit of each plant was also identified. These communities were compared to each other with a Sorenson Qualitative Index. The data sets were then separated by the woody species of each group and were compared to a list of tree species (Appendix A) that Bill Haber identified in the Los Llanos area (Figure 2). A Sorenson Qualitative Index was used to qualify these comparisons. RESULTS Seed Germination Plots For both of the seed germination species, a significant difference was found between treatments ( O. whitei 2 = 259.5 df = 2, and C. costaricensis 2 = 167 df = 2). For the O. whitei seeds, more seeds germinated in the infected plot than in the control and even more seeds germinated in the infected soil with the treatment of Bocconia seeds (Table 1). The C. costaricensis seeds were more difficult to analyze due to the loss of the control plot to a seed predator. Literature on C. costaricensis cites a 90% germ ination rate (Castro 1998), which was used for comparisons. However, based on observations of only three germinated seeds in the control before the attack (three days before collection), a 90% germination rate seems to be a far overestimate. Consequentl y, even though the numbers show significantly less germination in the infected area, the actual results probably would not have shown a significant difference. The germination rate in the Bocconia plot was similar to that of the infected. Infected Sapling s The fungus infected both species of saplings; however, no significant difference was found for the rates of infection between the two species ( 2 = 2.48 , df = 1). These results are very noteworthy given that 11 of the 14 C. costaricensis plants and 8 of the 13 plants were visibly infected after only three weeks time (Table 2). Evidence of Trichoderma
6 Evidence of Trichoderma was found in all sites with the highest abundance found in the infected area (Figure 3). Preliminary evidenc e showed that only one species of Trichoderma was present in all samples; however, the identity of this species and the presence and abundance of others has not yet been confirmed. Understory Plant Community Finally, results for the understory plant comp osition indicated that the regenerating infected area is similar to that of a pristine area (Appendix B). When comparing species of the family Lauraceae, often used as an indicator of the health of woody species in the area, all species present in the pri stine area were also in the infected. Furthermore, three Lauraceous species present in the infected area do not occur in the pristine. The Sorenson Quantitative Index showed that the infected and pristine area are both very similar to the tree data colle cted by Bill Haber (S tree data vs. infected = 0.34, S tree data vs. pristine = 0.35) and that the two areas are more similar to each other than to the tree data (S infected vs. pristine = 0.42). DISCUSSION The seed germination results indicate that Rosellinia does not affect seed germination and suggest that Bocconia may somehow improve conditions for O. whitei germination. However, the possibility of Bocconia as a biological control still remains uncertain. The germination rate of the C. costarice nsis was not affected by the presence of Bocconia , indicating that Bocconia may be species specific. Furthermore, considering that Rosellinia does not affect seed germination, the effects of Bocconia were not convincingly tested for its uses against the f ungus. The results for the infected saplings demonstrate that the fungus does not discriminate between the species it infects. Also, the results show that Rosellinia is not size or age specific. The results for Trichoderma disprove the theory that the absence of Trichoderma allowed Rosellinia to establish and are discouraging when considering Trichoderma as a biological control. Furthermore, if only one species of Trichoderma is present in all samples, and this species is not limiting Rosellinia , the entire area may be at risk for infection. The results indicate that the use of Trichoderma as a biological control may require the introduction of an exotic species. The understory plant community data show that the current seed bank in the infected area i s similar to that of a pristine. These results indicate that the regenerating forest will effectively replace the losses to Rosellinia if these species make it to adulthood and reproduce. However, if a biological control cannot be found for Rosellinia th ese plants will most likely be killed later in life and the forest will remain fragmented. These results outline a general model for Rosellinia in the infected area. Rosellinia does not affect seed germination; however, Rosellinia can infect very young pla nts within a time span of three weeks. This model indicates that the pathogen is a root specialist. Instead of forming a specific relationship with a certain species, Rosellinia is a general specialist for all root types. This conclusion is supported in many ways. First of
7 all, Rosellinia does not affect germination because germinating seeds lack roots; small understory plants are also seemingly unaffected at young ages because root material is not largely distributed throughout the infected soil. Howev er, once a plant reaches the sapling or adult tree stage it is at a greater risk for infection. This is very logical because as a tree matures and grows it also increases its root mass, increasing the chances for infection simply due to having more roots in the soil. Studies of Rosellinia in agriculture also provide support because Rosellinia most strongly affects monoculture crops (Alexopolos, et. al. 1996). This fits the model because if an area is made up of all the same species, Rosellinia does not h ave to adapt to any changes in the host plants available, it can simply spread without limit. Finally, the method Rosellinia uses to spread through root material also supports this conclusion. The spores of the fungus are specialized to travel through di rt and filamentous root material to infect the next plant (Petrini 1993). Unlike most fungi species that use wind or water to disperse, Rosellinia is adapted specifically for life underground as a general root specialist. A future study evaluating the e ffects of Rosellinia on taproots vs. filamentous roots would provide further evidence for this theory and strengthen the model. Unfortunately, this model still leaves gaping holes in our knowledge of Rosellinia , especially in how the pathogen originally e stablished in the Los Llanos area. If the species of Trichoderma found in the infected area is not a limiting factor, what is? Many possibilities remain. Possibly another species of Trichoderma was present in the area and is now, for reasons other than the presence of Rosellinia , absent allowing Rosellinia to dominate. Investigations of Trichoderma in other areas of the Monteverde region could examine this possibility. Other possibilities could be related to abiotic factors. For example, global climat e change could have somehow changed the community or conditions of the area making it more ideal for Rosellinia . More investigation is necessary on the exact temperature and moisture conditions that Rosellinia requires. Understanding why Rosellinia is pr esent in the area could be key to generating a biological control. The greatest need in the battle against Rosellinia is for interest and creativity. Traditional control methods have proved to be ineffective; consequently, we need more minds working to gether to combat this threat to the Monteverde Corridor. Given the current understory composition, the forest is ready to regenerate itself to its original state if the fungus can be stopped. With a little more time and creativity, Rosellinia can turn in to a success story for the Monteverde region instead of a tragedy. ACKNOWLEDGEMENTS Fist and foremost I must thank Karen Masters for the long hours spent discussing this project with me, her help analyzing my results, the endless phone conversations with INTA, and all the little things she did to help along the way. I would like to thank Leonel Arguedas for the use of his property, his assistance in identifying the fungus, and his trust in me to not further spread the pathogen. Thanks to Dr. Josslyn for his assistance in my search for Trichoderma . Thanks to the CIEE teaching assistants and Javier for helping me to find supplies. I would also like to thank EAP Jason and Sarah Stinson for their
8 advice and background information about Rosellinia . Finally , I must thank Tico Jason for the project idea, because without him I never would have began this project. LITERATURE CITED ALEXOPOULOS, C.W., BLACKWELL, M., C.J. MIMS 1996. Introductory Mycology . John Whiley & Sons: New York, NY. pp. 352, 356, 18 2, 243. ANONYMOUS 2003. Rosellinia Necatrix . < www.pyrenomycetes.free.fr/rosellina/html/Rosellinia_necatriz.htm> Acessed 2005 April 10. CASTRO, KAREN L YNNE 1998. Propagation manual for the trees of Monteverde, Costa Rica. In "Germination studies on two tropical tree species, Alnus acuminata spp. arguta (Schlechtendal) Furlow and Pithecellobium saman (Jacq.) Benth." University of Alberta, Edmont on. HABER, W.A., BELLO, E.,, ZUCHOWSKI, W 2000. An Introduction to Cloud Forest Trees. Mountain Gem Publications: Puntarenas, Costa Rica. pp. 93 and 157. HARMAN, G.E. 2004. Trichoderma spp. Deuteromyctes, Moniliaes. < www.nyases.cornell.edu/ent/bi ocontrol/pathogens/trichoderma.htm> P P R R I I L L L L , , 2 2 0 0 0 0 5 5 . . H H Y Y P P P P P P a a PRILL, 2005. HYPP Pathology Homepage < www.inra.fr/Internet/Products/HYP3/Pathogene/6/rosnee.html> Ace ssed 2005 April 10. MASTERS, K. Personal Communication. April 15, 2005. PETRINI, L. E. 1993. Rosellinia species of the temperate zone. Sydowia 44: 169 281. C C o o s s t t a a STINSON, S. 2004. Fungal distribution of Rosellinia species in Monteverde, Cos ta Rica: an analysis of horses as possible vectors in secondary forest fragments. EAP, Fall 2004.
9 FIGURE 1: Appearance of the mycelium of Rosellinia on an infected root. (Photo by Jessica Smith) FIGURE 2: Plot of the Los Llanos are a used to generate tree composition data. Information from transects 44, 54, 57 and 58 were used to compare the original tree composition to the regenerating woody species in the infected area (map by Bill Haber).
10 Trail Wet Site A Site B Site C Site D FIGURE 3. Cultures of Trichoderma sp . yellow colonies correspond to the presence of the fungus. Trail wet and site A correspond to the infected area confirmed to have Rosellinia , site B corresponds to the area suspected to have Rosellinia and Sit es C and D correspond to the pristine area. Highest abundance of Trichoderma found in the infected area, lowest/moderate amounts found in the pristine area.
11 TABLE 1. Number of seeds germinated, not germinated, and germination rate for Ocote a whitei and Citharexylum costaricensis for three treatments: control, infected, and infected plus Bocconia. Results indicate that Rosellinia does not deter germination and that Bocconia may improve conditions for germination. C. costaricensis control da ta lost, 90% germination rate from literature thought to be an overestimate. treatment # seeds germinated # seeds not germinated germination rate Ocotea whitei control 38 60 .39 infected 72 20 .78 inf + Bocconia 86 14 .86 Citharexylum costaricensis control .90 infected 19 82 .19 inf + Bocconia 20 90 .18 TABLE 2. Number of and Citharexylum costaricensis individuals showing visual signs of infection and infection rates by Rosellinia after three we eks. No significant difference found between the two species. Citharexylum costaricensis Infected 11 8 Not infected 3 5 Infection rate .79 .61
12 APPENDIX A. Tree species data collected by Bill Haber for the Los Llanos area of Monteverde Costa Rica . Beilschmiedia brenesii Beilschmiedia chancho blanco Billia colombiana Bourreria costaricensis Bunchosia veluticarpa Calyptranthes pallens Cedrela tonduzii Cinnamomum brenesii Croton mexicanus Cupania glabra Dendropanax arboreus Diospyros hartmanniana Eugenia guatemalensis Ficus laterisyce Ficus tuerckheimii Hasseltia floribunda Inga mortoniana Inga punctata Matayba oppositifolia Meliosma idiopoda Nectandra sali cina Ocotea floribunda Ocotea llanos Ocotea whitei Ormosia cruenta Persea amsl Pouteria exfoliata Pouteria reticulata Sideroxylon portoricense Sorocea trophoides Stauranthus perforatus Styphnolobium monteviridis Tapirira mexicana Trophis racemosa Xylosma chlorantha
13 APPENDIX B. Species collected from the infected and pristine area of Los Llanos, Monteverde. Family Genus species area found habit Araceae Syngonium sp both vine Arailiaceae Dendropanax arboreus both woody Arecaceae Chamaedorea costaricana both herb Asclepidaceae Gonolobus edulis both vine Asclpiadaceae Cynanchum filisepalum both vine Asteraceae Vernonia arboresence both liana Bignoniaceae Martinella obovata both liana Cucurbitaceae Melothria pen dula both vine Fabaceae Rhynchosia edulis both liana Hippocrateaceae Billia colombiana both woody Lauraceae Ocotea "los llanos" both woody Lauraceae Nectandra salicina both woody Malvaceae Sida rhombifolia both herb Mimosaceae Inga punctata both wood y Mimosaceae Zygia palmana both woody Mimosaceae Inga mortoniana both woody Mimosaceae Inga marginata both woody Moraceace Sorocea trophoides both woody Myrtaceae Eugenia guatemalensis both woody Myrtaceae Eugenia monticola both woody Olacaceae Heis teria acuminata both herb Piperaceae Piper amalago both herb Poaceae Oplismenus burmannii both herb Pteridaceae` Pteris sp both herb Rosaceae Rubus roseifolius both herb Rubiaceae Psychotria quinqueradiata both herb Rutaceae Zanthoxylum fagara both w oody Sabiaceae Meliosma idiopoda both woody Solanaceae Solanum brenesii both woody Solanaceae Solanum cordovense both woody Amaranthaceae Pleuropetalum sprucei infected herb Amaranthaceae Iresine diffusa infected herb Apocynaceae Forsteronia montever densis infected liana Asteraceae Conyza bonariensis infected herb Asteraceae Archibaccharis panaensis infected liana Asteraceae Conyza canadensis infected herb Cecropiaceae Cecropia obtusifolia infected woody Commelinaceae Tradescantia zanonia infecte d herb Cucurbitaceae Gurania sp infected vine Cucurbitaceae Sicydium tamnifolium infected vine Euphorbiaceae Sapium glandulosum infected woody Euphorbiaceae Euphorbia elata infected woody Lauraceae Persea "small leaf" infected woody Lauraceae Beilsch miedia "choncho blanco" infected woody Lauraceae Nectandra membranacea infected woody Malvaceae Pavonia rosea infected herb Malvaceae Hampea appendiculata infected woody Marantaceae Calathea crotalifera infected herb Marantaceae Calathea insignis infe cted herb
14 Myrantaceae Stromanthe tonckat infected herb Orchidaceae Corymborkis forcipigera infected herb Papaveraceae Bocconia frutesens infected woody Phytolaccaceae Phytolacca icosandra infected herb Phytolaccaceae Rivina humilis infected herb Pipe raceae Piper auritum infected woody Piperaceae Piper dotanum infected herb Piperaceae peperomia sp infected epiphyte Piperaceae Piper hispidum infected herb Rutaceae Stauranthus perforatus infected woody Sapindaceae Paullinia costaricensis infected l iana Sapotaceae Pouteria exfoliata infected woody Solanaceae Cestrum racemosum infected woody Solanaceae Cestrum megalophyllum infected woody Solanaceae Solanum umbellatum infected woody Solanaceae Witheringia solanacea infected herb Solanaceae Solan um americanum infected herb Solanaceae Solanum tuerckheimii infected herb Solanaceae Solanum capsicoides infected herb Solanaceae Solanum chrysotrichum infected woody Solanaceae Cestrum "lace leaf" infected herb Solanaceae Brachistus stramoniifolius i nfected herb Ulmanaceae Trema micrantha infected woody Urticaceae Pouzolia phenacoides infected woody Acanthaceae Justicia valerii non herb Apocynaceae Mandevilla veraguasensis non liana Araceae Philodendron anisotomum non vine Arecaceae Chamaedorea parvifolia non herb Asteraceae Ageratum sp non herb Asteraceae Mikania sp non vine Asteraceae Clibadium surinamense non herb Asteraceae Koanophyllon hylonomum non woody Asteraceae Elephantopus mollis non herb Asteraceae Baccharis trinervis non liana Blechnaceae Blechnum occidentale non woody Convolvulaceae Ipomoea batatas non vine Erithoxilaceae Erithoxlum macrophillum non woody Fabaceae Ormosia cruenta non liana Flacourtiaceae Xylosoma chlorantha non woody Lamiaceae Salvia costaricensis non her b Malpighiaceae Tetrapteris schiedeana non liana Meliaceae Guarea "dwarf" non woody Menispermaceae Abuta panaensis non liana Mimosaceae Inga sp non woody Mimosaceae Cojoba valerioi non woody Myrsinaceae Ardisia compressa non woody Myrtaceae Psidium guaiava non woody Myrtaceae Myrcia splendens non woody Myrtaceae sp sp non woody Nytaginaceae Neea psychotrioides non woody Nytaginaceae Pisonia sylvatica non woody Piperaceae Piper bredemeyeri non herb Piperaceae peperomia sp #2 non epiphyte Poace ae sp non herb
15 Rhamnaceae Gouania polygama non liana Rubiaceae Hamelia patens non woody Rubiaceae Chiococca "long petiole" non woody Rubiaceae Chiococca phaenostemon non liana Rubiaceae Psychotria pubescens non herb Sapindaceae Exothea paniculata no n woody Sapindaceae Paullinia cururu non liana Sapindaceae Cupania glabra non woody Sapotaceae Pouteria exfoliata non woody Schizaeaceae Anemia sp non herb Similaceae Smilax spinosa non liana Thymeliaceae Daphnopsis americana non woody