Spider abundance and diversity in Life Monteverde Flores 1 Comparison of s pider abundance and diversity in different habitats in the Life Monteverde coffee farm Kristen Flores Department of Environmental Science, Policy, and Management University of California, Berkeley EAP Fall 2016 16 December 2016 ABSTRACT Coffee agroecosystems have abundant insect prey, and therefore have the potential to supp ort a high diversity of spiders. I n this study, I measured spider diversity in the Life Monteverde low shade coffee farm located in Caitas Guanacaste, Costa Rica. I compared the alpha and beta diversity of an isolated coffee field a forest surrounded and more highly shaded coffee field, and a forest patch all located on the farm. At each site, I coll ected spiders from 12 plots, and later I identified each spider to the family level. In total, I collected 508 individuals of 15 families. Using the true diversity index, I found that the isolated field had the highest family richness, while the forest patch had the highest alpha diversity when considering evenness of family distribution. The isolated field and the forest surrounded field had the most family ove rlap according to the Jaccard index which measures beta diversity Theridiidae was the most common family found in each site, and the second most c ommon family varied among sites These results suggest that while low shade coffee growing can support a high diversity and abundance of spiders, populations may be dominated by a few families. The high family complementarity between the two most similar s ites, the field s, indicate that habitat composition, rather than proximity to forest, has a greater e ffect on the presence of specific spider families Comparacin de abundancia y diversidad de araas en diferentes hbitats de la finca de caf Life Monteverde RESUMEN Los agroecosistemas de caf presentan una abundancia de insectos muy variada, y por eso tienen el potencial de soportar una alta diversidad de araas. En este estudio, yo med la diversidad de araas en la finca de caf Life Monteverd e, ubicada en Caitas, Guanacaste, Costa Rica. Compar la diversidad alpha y beta entre un campo de caf aislado, un campo de caf rodeado de bosque que contiene ms rboles, y un terreno de bosque, todos dentro de la finca. En cada sitio, colect araas e n 12 parcelas, y luego las identifiqu a nivel de familia. En total, colect 508 individuos de 15 familias. Segn el ndice Diversidad Verdadera, el campo aislado tena el nivel ms alto de familias diferentes, mientras que el terreno de bosque tena la di versidad ms alta considerando la uniformidad de la distubucin de las familias. Los dos campos de caf tenan ms similaridad de familias segn el ndice Jaccard, que mide la diversidad beta. Theridiidae fue la familia ms comn en cada sitio, y la segund a familia ms comn vari entre sitios. Estos resultados sugieren que aunque las fincas de caf con menos sombra pueden soportar una alta diversidad y abundancia de araas, quizs las poblaciones son dominadas por pocas familias. La alta complementaridad d e las familias entre los dos sitios ms similares que corresponden a los dos campos de caf, indica que la composicin de hbitats tiene una consecuencia ms importante que la proximidad del bosque en cuanto a la presencia de familias especficas de araas
Spider abundance and diversity in Life Monteverde Flores 2 Agricultural ecosystems can support a wide diversity and abundance of spiders (Araneae) (Barrera 2008). In such settings, spiders can perform beneficial services by eating insect pests such as flies, katydids and beetles that would o therwise damage cro ps (Foelix 2011). Coffee plantations may especially benefit from the presence of spiders, as insects are considered to be the most prominent coffee pests (Barrera 2008). For example, some spiders in Costa Rica consume long horned grasshoppers and broca beetles, which destroy coffee cherries (Barrera 2008). Thus, coffee plantations can support spiders by providing them ample food sources; in turn, spiders act as a natural pest control. Spiders are additionally necessary in agroecosystems as a prey item Coffee p ollinators such as wasps, birds, and bats fee d upon spiders (Foelix 2011). Higher abundances of spiders could attract pollinators to coffee farms, thereby augmenting yields. Despite their role as predators, spiders are typically only considered si gnificant to insect population control when present in high abundance (Pinkus Rendn et al. 2006). Because of this, maintaining high spider population sizes can substantially decrease insect pest population size. Since different spiders consume different t ypes of prey, a higher diversity of spiders on farms also acts as a more comprehensive pest control. Coffee plantations in Costa Rica have the potential to support high spider abundance and diversity by maintaining high habitat complexity. In the traditional method of coffee farming, coffee is grown under shade trees of up to forty different tree species (Perfecto et al. 1996). Shade tree farming creates considerable structural diversity that can serve as suitable habitat for a high diversity of spider families (Perfecto et al. 1996). However shade coffee is increasingly being replaced by sun coffee, in which all trees are removed and coffee is grown in a monoculture (Jha et al. 2014). The trend towards homogenization may represent a threat to spiders because spiders are particular in their specific temperature, humidity, wind, and light intensity requirements (Foelix 2011). In forests, different species of spiders are distributed in layers classified by height from the ground (Foelix 2011 ). The availability of a variety of these conditions decreases with habitat homogenization. Therefore more complex habitats, such as forest s and shade tree fields, are expected to support higher spider diversity Spider diversity may also be affecte d by interconnectivity among habitats. Spide rs that disperse by ballooning, the process of releasing threads th at are carried away by the wind, typically only travel a few meters (Foelix 2011 ). Other spiders, such as tarantulas and wandering spiders, disperse by walking (Foelix 2011 ). So, it is more likely that spiders will establish successfully if there is suitable habitat nearby (Foelix 2011 ). Due to the nature of spider dispersal mechanisms, habitat sites that are closer together likely have a grea ter intermixing of populations, and therefore similar family composition. Although previous studies have reported th at spider diversity increases with habitat complexity, this finding has yet to be heavily supported in agricultural landscapes. In some ca ses, spider diversity is dramatically lower in crop fields than natural environments (Foelix 2011, Banks et al. 2007). For instance, Banks et al. (2007) showed that spider diversity increased considerably in forest fragments compared to fruit and coffee mo nocultures. However, other studies have found that spider diversity is higher in certain agricultural settings than undisturbed
Spider abundance and diversity in Life Monteverde Flores 3 environments (Banks et al. 2007). For example, Pinkus Rendn et al. (2006) found that species richness of spiders increased in l ow shade management compared to less homogenous coffee systems. In contrast, Marn and Perfecto (2012) concluded that removing shade trees does not significantly affect spider diversity within fields. Due to the various outcomes of studies performed on this topic, the purpose of this study is to answer the question, H ow do spider abundance and diversity change among different answer this question, I compared spider diversity of a coffee field is olated from forest, a coffee field with higher abundance of shade trees and that is surrounded by forest, and a secondary forest patch The goal of this question is to more clearly understand how spider diversity and abundance are affect ed by different hab itats within this specific coffee agroecosystem I expect that spider diversity and abundance will increase with habitat complexity; the forest will be the most diverse habitat, while the isolated field will be the least. I also expect that the forest and the forest surrounded field will have the highest number of families in common because they are located the closest to each other and therefore spiders will be more easily able to move between these two sites. MATERIALS AND METHODS Study Sites My study took place in the Life Monteverde coffee plantation located in Caitas, Guanacaste (10.324842, 84.843076). The plantation is representative of a low shade coffee system. For most fields, n ative woody trees and banana plants are present in low abu ndances within rows of coffee trees. Coffee trees are planted in rows cleared of other vegetation, although low grasses were often present in each field Each field is encircled by a windbreak comprised of planted non native cypress trees as well as native trees. I sampled three different sites within the farm (Figure 1) The first site (Site A : Isolated Field ) is a coffee field of approximately 3, 300 m This field is bordered by windbreaks from the North, South, and East. Beyond these windbreaks are more coffee fields. The western border of the field is a public road with cleared land beyond it There are about 10 trees planted in this field. Compared to other coffee fields on the farm Site A is relatively distant from any forest patch. Thus, Site A represents a n isolated field with low shade tree coverage Site B (Forest Surrounded Field) is another coffee field that is approximately 3,600 m Its w estern border is a forest patch and the remaining borders are windbreaks. However, beyond each windbreak border are small coffee fields followed by forest. Within the field, there are over 40 individual tree s of 8 different species. Site B represents a forest surrounded, more highly sha ded field. Site C (Forest Patch) is a forest patch within the farm. It is approximately 40,000 m The forest patch is secondary forest that has been growing for less than 30 years, and it also has a creek running through it. Site C represents natural or undisturbed environment.
Spider abundance and diversity in Life Monteverde Flores 4 Figure 1: A map of the location of each study site (A, B, C) on Life Monteverde. Each boxed in area is a coffee field, while the darker areas represent forest. Methodology My sampling method s consisted of using transects for each site (Figure 2) Sites A, B, and C each have paths in the center of either site ; t his path acted as a main transect. Starting from the beginning of each site I walked 15 meters down this transect. At this point, I walked 3 meters perpendicular to the transect and took a sample. I did this two times in each direction that was perpendicular to the main transect. I then continue d down the main transect for 15 meters again and repeat ed the process two more times, for a total of three horizontal transect samples, each with four total sample sites. In total, each site was sampled 1 2 times in various locations. I used this sampling method at each site in order to look at the sites overall, and not have bias in terms of edge effects or proximity to the main path. I sampl ed for five days at the farm 15 Nov, 17 Nov, 21 Nov, 22 Nov, an d 23 Nov, 2016. I sampled every site eac h day and alternated which sites I sampled first every day to avoid bias based on time of day. I collected samples of spiders using a beating sheet. At each specified plot, I placed an open umbrella upside down on the ground. Then, I shook the surrou nding plants for thirty seconds, which caused the spiders to fall into the umbrella. Each plot was 1 m by 1 m, and I beat up to 2 m in height. With a pair of tweezers, I collected all spiders that had fallen into the A B C A : Isolated Field B: Forest Surro u nded Field C : Forest Patch
Spider abundance and diversity in Life Monteverde Flores 5 umbrella. I also collected any spiders that I noticed remained in my plot but did not fall after beating. I placed all spiders into a plastic container filled with ethanol for later identification. After I completed my collection, I identified each individual to the family level using the guide Spiders of North America (Ubick et al. 2005 ) Figure 2: Transect method used for each site. Each dot represents a sampling plot. In total, 12 plots were sampled at each site per day. Data Analysis To measure the alpha diversity of each site, I used the True Diversity index : q D= ( P i q ) 1/1 q Where q is an order of diversity that can range from [0, and q D is number of species Q =0 measures family richness. and weigh t s each family according to their relative abundance. Q=2 is the inverse Simpson concentratio n and weights dominant families To measure the family turnover between each site (beta diversity), I used t he Jaccard index of similarity: C= J/(a+b J)
Spider abundance and diversity in Life Monteverde Flores 6 Where J= the number of families shared between two sites, and a and b are the total number of families in each respective site. This index ranges from 0 1, with higher values indicating more similarity bet ween sites. RESULTS Alpha Diversity I collected a total of 508 individual spider s. The forest surrounded field had the highest abundance of individuals with 212 spiders collected the isolated field had 152 individuals, and the forest patch had the lowest with 144 spiders collected Theridiidae was th e most dominant family in all three sites In the two field sites Salticidae was the second most common family; for the forest patch Tetragnathidae was the second most common family ( Figure 3 ) According to the true diversity index ( Table 2 ), the isolated field had the highest family richness (q=0) and the forest surrounded field had the lowest. When families were weighted according to their relative abundances (q=1), the forest patch had th e highest diversity, and the forest surrounded field had the lowest. When dominant families were weighted more heavily (q=2), the forest was the most diverse and the forest surrounded field was the least. Table 1: A list of the number of individuals per family found at each site. Family Isolated Field (A) Forest Surrounded Field (B) Forest Patch (C) Anyphaenidae 14 5 0 Araneidae 8 16 32 Corinnidae 1 1 0 Dictynidae 1 0 0 Lycosidae 2 0 0 Pisauridae 0 0 3 Pholcidae 0 0 1 Salticidae 47 64 20 Scytodidiae 0 0 2 Sparassidae 5 0 1 Tetragnathidae 10 32 36 Theridiidae 61 84 44 Thomisidae 1 3 2 Uloboridae 1 7 3 Unidentified Juvenile 1 0 0 Total Individuals 152 212 144 Total Families 12 8 10
Spider abundance and diversity in Life Monteverde Flores 7 Figure 3: The most common families found at each site and their abundances For Site A, Others includes the families Corinnidae, Dictynidae, Lycosidae, Sparassidae, Thomisidae, Uloboridae, and unidentified juvenile. For Site B, Others includes the families Corinnid ae, Thomisidae, and Uloboridae. For Site C, Others includes the families Pisauridae, Pholcidae, Scytodidiae, Sparassidae, Thomisidae, and Uloboridae Table 2: True diversity of spiders of each study site, with three different orders of diversity measured Q=0 measures family richness, q=1 measures diversity by weighting families in proportion to their abundance, and q=2 measures diversity by weighting dominant families. True Diversity q D Site A : Isolated Field Site B : Forest Surrounded Field Site C : Forest Patch q=0 12 8 10 q=1 5.04 4.46 5.29 q=2 3.65 3.59 4.43 61 84 44 47 64 20 10 32 36 8 16 32 14 5 0 12 11 12 0 10 20 30 40 50 60 70 80 90 Isolated Field (A) Forest-Surronded Field (B) Forest Patch (C) Number of Individuals Families Theridiidae Salticidae Tetragnathidae Araneidae Anyphaenidae Others
Spider abundance and diversity in Life Monteverde Flores 8 Beta Diversity I collected a total of 15 different families The isolated field had the highest number of families with 12, the forest patch had 10 families, and the forest surrounded field had 8 families ( Figure 4 ). Both the isolated field and the forest patch had three families unique to each of those sites while the forest surrounded field had none. According to the Jaccard Index, the two field coffee sites were th e most similar because they shared the most families (C=0.667) The isolated coffee field and the forest patch have the highest family turnover rate (C=0. 467) which means that families changed the most between those sites Figure 3 shows the overlap of fa milies between each site. Figure 4 : Family overlap between each site C symbolizes the complementarity between each site according to the Jaccard index. DISCUSSION I expected that the forest patch would have the highest diversity because has the greates t habitat variation and complexity. However, I found that the forest patch had the lowest number of families and individuals. The isolated field, which is the most homogenous site had the highest number of families while the forest surrounded field had the most individuals Similarly, Pinkus Rendn et al. (2006) found that alpha diversity of spider species was higher in low shade coffee fields than rainforest patches in some cases. This may be attributed to the difference s in the way that vegetation is structured in forest versus field. Although the fields are low shade monocultures, they are densely vegetated with coffee plants. Each coffee plant could support many spiders often, I collected around ten spiders per plot in the field sites. In contrast, the forest is less densely vegetated. Sometimes, my plots in the forest only contained one overhanging stem, and as a result I did not collect any spiders in a plot on several occasions.
Spider abundance and diversity in Life Monteverde Flores 9 Although fields are less structurall y complex, they contain more available habitat. Additionally, coffee attracts an abundance of insects when flowering or fruiting, which can support a higher abundance of spiders. Because of this spiders can establish in higher abundance in the fields than in the forest Despite its lower richness and abundance, the forest patch is the most diverse site when evenness is considered (q=1 and q=2) In my study, t he homogeny of the fields allows the spider families that prefer dense vegetation to dominate; other families are present but in significantly lower abundances (Figure 3). In contrast, there is more variety of habitat types in the forest and consequently f ewer dominant families. This result is supported by Perfecto et al. (1996) findings that arthropod diversity increases when coff ee fields most resemble forests. This pattern is based on the idea that the more vegetation diversity, the more spider diversity an area can sustain. Unexpectedly, the isolated, less shaded field was more diverse than the fores t surrounded field. Other studies have similarly found that spider diversity increases with land management such as more homogenous coffee fields (Pinkus R endn et al. 2006; Marn and Perfecto 2012). The lower diversity of the forest surrounded, more highly shaded field may be attributed to a More shaded coffee systems support higher bird abundance (Perfecto et al. 1996), which may deplete spider diversity. The isolated field and the forest surrounded field share the highest family complementarity, while the forest patch and the isolated field sha re the least. The two fields likely have the highest complementarity because they are very similar habitats and consequently can support similar families The forest and the isolated field possibly shared the fewest families because of the stark difference in habitat type and the large distance between the sites. This large distance would likely limit the rate of dispersal that occurs between these two sites In comparison, the forest surrounded field is adjacent to the forest, enabling dispersal and result ing in family overlap. Still, the proximity of the forest may contribute to the diversity of both fields, as previous studies have shown that spider diversity in agriculture increases with proximity to forest patches (Banks et al. 2007). Theridiidae was the most abund ant family found in every site. This is probably because they are one of the largest families of spiders, with more than 100 genera ( Ubick et al. 2005). Similarly, Salticidae is one of the most common spider families in the tropics (Ubic k et al. 2005), and it was the second most abundant family in both of the field sites. Tetragnathidae was the second most abun dant family in the forest site. This is likely due to their affinity towards constructing webs near water as the forest site has a creek running through it (Foelix 2011). Since I only identified spiders to the family level, future studies could achieve a more precise understanding of diversity by identifying spiders to the genus or species level. This would allow for more specific d iversity comparisons. In this study, I only sampled spiders that were below 2 meters in height; however, the forest contained more trees and tall plants than the field sites. Future studies should take into account upper habitat strata to obtain a more complete understanding of spider diversity. It is probable that the forest contains more diversity higher up in the trees. It would also be interesting to incorporate samples of ground dwelling spiders in the soil strata of all habitat types.
Spider abundance and diversity in Life Monteverde Flores 10 Addit ionally, my study does not encompass differences in spider diversity throughout the year. Pinkus Rendn (2006) found that spider species composition changed between seasons, so it would be useful to study how diversi ty varies throughout the year for each h abitat type. This study demonstrates that coffee agriculture can support considerable spider diversity and abundance Even the most homogenous site of my study, the isolated field, had similar diversity values to the forest patch. The maintenance of undisturbed habitats on the farm, such as t he windbreaks and forest, can serve as areas for spiders to establish populations (Foelix 2011) and likely contribute to the high diversity of Life Monteverde. As one of the most efficient predators of inverte brates (Ubick et al. 2005), spiders represent a critical component of coffee agroecosystems, and can be considered to be effective agents of biocontrol. ACKNOWLEDGEMENTS I would like to acknowledge my gratitude for my advisor Emilia Triana Her patient guidance, endless help, and passion for spiders made it possible for me to complete my project and enjoy doing so. I would also like to thank Sofa Arce Flores for her feedback and assistance, as well as all the other EAP staff members for su pporting me along the way. Thanks to the Instituto Monteverde for allowing me to use their lab space. I am grateful to the staff at Life Monteverde, especially Guillermo Vargas and Daniel Vargas, for allowing me to conduct my research on their farm and for always going out of their way to assist me. Thanks to my fellow Finca Fam and Lab Crew students for providing motivation and entertainment throughout the project. Finally, I would like to thank my parents for making this experience possible for me. LIT ERATURE CITED Banks, John E., P. Sandvik, and L. Keesecker. 2007. Beetle (Coleoptera) and spider (Araneae) diversity in a mosaic of farmland, edge, and tropical forest habitats in Western Costa Rica. The Pan Pacific Entomologist 82: 152 160. Barrera, Juan F. 2008. Coffee pests and their management. Encyclopedia of Entomology 961 998. Ubick, D., P. Paquin, P.E. Cushing, and V. Roth (eds). Spiders of North America: an identification manual American Arachnological Society, 2005. Foelix, Rai ner F. Biology of Spiders 3 rd ed., Oxford University Press, 2011. Jha, Shalene, C.M. Bacon, S.M. Philpott, V.E. Mndez, P. Lderach, and R.A. Rice. 2014. Shade coffee: an update on a disappearing refuge for biodiversity. BioScience 64: 416 428. Marn, Linda, and Ivette Perfecto. 2012. Spider diversity in coffee agroecosystems: the influence of agricultural intensification and aggressive ants. Entomological Society of America 42: 204 213. Perfecto, Ivette, R.A. Rice, R. Greenberg, and M.E. va n der Voort. 1996. Shade coffee: a disappearing refuge for biodiversity. BioScience 46: 598 608. Pinkus Rendn, Miguel Angel, G. Ibarra Nez, V. Parra Tabla, J.A. Garca Ballinas, and Y. Hnaut. 2006. Spider diversity in coffee plantations with diff erent management in Southeast Mexico. The Journal of Arachnology 34: 104 112.