Insect Communities in Fragments Laine 1 Insect communities across forest fragments of different sizes in Monteverde Jesse Laine EAP Fall 2017 University of California Santa Cruz 15 December, 2017 ABSTRACT Forest fragmentation creates numerous boundaries and microclimates that affect organisms of all taxa. I studied the composition of insect communities across two forest fragments of different sizes, as well as continuous forest and pasture. Pitfall traps an d sweep netting techniques were used to sample 888 insects overall. The fragments had similar numbers of insects and morphospecies, regardless of size. The pasture had the most morphospecies present. In addition, Forest edges consistently showed higher abundance and diversity than forest cores. This indicates that reduction of habitable land does not always reduce the numbers of individuals and species, and in some cases can increase them. Only four of the total 145 morphospecies sampled were found living in every site. This means 2.8% of morphospecies were resilient and generalist enough to exist across the different habitats. This research shows that the differences in habitat created by fragmentation are exceptionall y difficult for insects to man age Fut ure research should do similar samples in different regions and climates. Comunidades de insectos en fragmentos de b osques de d iferentes t amaos en Monteverde RESUMEN La fragmentacin del bosque crea numerosos lmites y microclimas que afectan a organism os de todos los taxones. Estudi la composicin de las comunidades de insectos en dos fragmentos de bosques de dife rentes tamaos, as como un bosque continuo y un pastizal Utilic dos mtodos para colectar insectos: trampas de cada (pitfall) y barrido con red de mariposas, y colect 888 insectos en total Los fragmentos tenan nmeros similares de insectos y morfoespecies, independientemente del tamao. El pasto tena la mayor cantidad de morfoespecies presentes. Adems, los bordes del bosque mostraron consistentemente mayor abundancia y diversidad que los ncleos forestales. Esto indi ca que la reduccin de reas habitable s no siempre reduce el nmero de individuos y especies, y en algunos casos puede aumentarlos. Solo cuatro de las 145 morfoespecies mue streadas se encontraron en todos los sitios, lo cual significa que el 2.8% de las morfoespecies eran lo suficientemente resistentes y generalistas como para existir en los diferentes hbitats. Esta investigacin muestra que las barreras creadas por la frag mentacin son excepcion almente difciles de cruzar por insectos Futuras investigaciones deberan hacer muestras si milares en diferentes regiones y pocas de ao.
Insect Communities in Fragments Laine 2 Forest fragmentation is a widely occurring event where deforestation for huma n development leaves a once continuous forest in a broken, patchy state. This causes the forest to undergo several changes, including a shift in plant communities and microclimate. These fragments create barriers for some species and can limit their mobility between patches. Species are hindered from travelling across fragments by the distance between habitats, abiotic barriers such as roads that separate the potential living space, and the differing conditions within each fragment Fragmentation also creat es two smaller habitats within each fragment. Forest fragment edges have different environmental conditions, such as higher sun and wind exposure, and therefore have diff erent flora and fauna than core forest (Murcia 1995). Multiple studies show the effe ct of forest patches on tropical insect diversity and abundance (Klein 1989, Hill 2003). These studies often correlate decreasing fragment size with decreasing populations of spec ific insects (Jenkins 2013). Fewer studies have examined the relationships of several close patches of forest in th e context of insect communities and fragment size. Insect community studies suggest that there is not necessarily an overall species loss, but rather a species turnover in the composition of the community (Summerville 2004). When looking at several fragments, the overlap of species across sites can reveal the probability of certain species to travel across fragments and proliferate (Hill 2011). I sampled insects in order to characterize insect community composition acro ss several fragments and compare presence of morphospecies. This information allowed me to explore which insects are better able to travel and proliferate amongst divided patches of habitat. Knowing these species and the mechanisms behind their distributi on, is crucial to understanding the future of tropical insect communities. It is important to predict which insect groups could increase in presence as a result of fragmentation because insects exist at a low trophic level, and are a food and pollination s ource for a large amount of other tropical forest taxa. The layout of the insect community can also predict the future plant communities (Andresen 2003). This study explores how different fragment habitats a ffect insect communities, and which species are r esilient enough to exist across multiple different sites. MATERIALS AND METHODS I established four test sites in the Monteverde area of Puntarenas, Costa Rica. Two of these sites were small (~two hectares) and large (~4 hectares) forest fragments (Finca Arces and Finca San Francisco, respectively). The third site was a grassy pasture that separates the two forest fragments. The fourth site was a large section of continuous forest across a road from the three other sites, in a reserve called the Santuari o Ec logico (Figure 1).
Insect Communities in Fragments Laine 3 Figure 1 The four study sites. A is the large fragment (Finca San Francisco). B is the pasture. C is the small fragment (Finca Arce ). D is the continuous forest (Santuario Ecolgi co). For each site, I used pitfall traps and sweep netting to collect samples. In the small and large fragments, I set ten pitfall traps along the forest edge (within 15 meters of the tree line) spaced between five and ten meters apart. I then set ten pitfall trap in the forest core (~100 meters away from the edge) with similar spacing between traps. In the pastu re, 20 pitfall traps were set all along the edge, as the pasture core was not conducive to setting traps. In the continuous forest, 20 pitfall traps were placed along the trails running through the core of the forest. The pitfall traps all had an opening of ~5cm. Each trap was filled with ~2 centimeters of a mixture of 80% water, 20% ethanol, and several drops of soap. The pitfall traps were collected 24 hours after being buried. The contents were strained and the insects were separated from the leaves, soil, and crustaceans that had fallen into the traps. For sweep netting, I defined three different areas (~10x10 meters each) to sweep. In the small and large fragments, these three sites were locat ed on the edge, ~50 meters from the edge, and in the core. In the pasture, these sites were along the edge bordering the small fragment, the edge bordering the large fragment, and the core. In the continuous forest, these three sites were dispersed thro ugh the core. For each site at around 11:00am I would swing the butterfly net in a figure eight pattern while slowly moving along the 10x10 meter area. I would vary the height of the net as I swept along the vegetation. After sweeping the full area which usually took around 15 minutes I would place the mass of plant matter and insects that had accumulated in the net into a labeled plastic bag. I placed the sweep net samples in a freezer for at least 12 hours before handling. All sweep netting was done in close vicinity to the pitfall traps in the area. Sampling occurred between 12 Nov ember 2017 and 25 November 2017. I pinned all collected samples and separated them
Insect Communities in Fragments Laine 4 based on location and trap type. The collection was identified to order and sorted to morphospe cies using a decimal code, and all information was entered into Excel. I then compared morphospecies layouts amongst all sites, as well as examined data based on abundances. RESULTS In total, 888 insects were collected across seven orders and 145 morphospecies. The two fragments and the pasture had similar numbers of insects, while the continuous forest had significantly less (Figure 2a). Similarly, the fragments and pasture had similar numbers of morphospecies present, while the continuous forest had less (Figure 2b). The pasture had the most morphospecies present, despite having less individuals than the fragments. Diptera (flies) was the most prevalent order overall, but Hemiptera ( true bugs) was more abundant within the pasture (Figure 2c). The fragments and continuous forest follow similar patterns of order distribution, while the pasture has a different layout. Each site had a subset of isolated species that were only found at t hat specific site. The pasture had the most isolated species present, while the two fragments had similar levels of isolated species (Figure 2d). The pasture was the most diverse area, based on the reciprocal Simpsons index (22.0). The small fragment (18.7) was more diverse than both the large fragment (9.8) and the continuous forest (14.9). 0% 10% 20% 30% 40% 50% 60% 70% Small fragment Large fragment Pasture Continuous forestPercentage of Total InsectsSite Diptera Orthoptera Hymenoptera Coleoptera Hemiptera 62 61 69 28 0 10 20 30 40 50 60 70 80 Small fragment Large fragment Pasture Continuous forestNumber of MorphospeciesSite 0 50 100 150 200 250 300 Small fragment Large fragment Pasture Continuous forestNumber of InsectsSite (a) (b) (c)
Insect Communities in Fragments Laine 5 Figure 2 (a) the total number of insects across all sites. (b) the total number of morphospecies in each site. (c) the number of each insect orde r across all sites. (d) the number of isolated morphospecies for each site. There was a consistent difference between the communities of edge forest and core forest. Across both fragments, edge samples always contained more individual insects than core samples (Figure 3a). Additionally, edges (n=31) had higher numbers of isolated species than forest cores (n=20) (Figure 3b). Figure 3 (a) the total number of insects in edges and cores. (b) the number of morphospecies only found in edges and cores. In order to examine the permeability of the natural borders that form when forests are fragmented, I observed the overlap of species amongst all four sites. Only four morphospecies were found in all four sites, 2.8% of the total morphospecies ( Figure 4). In addition, only 12 species were found in three of the four possible sites. This means 11% of the total unique morphospecies were found in at least three sites simultaneously. Every other species was either isolated to one site, or was found in two of the four sites. The four species that were found in all sites were also highly concentrated within either the core of the forest fragments, or the core of the continuous forest (Figure 5). Three of the species 0 5 10 15 20 25 30 35 40 45 Small fragment Large fragment Pasture Continuous forestNumber of Isolated SpeciesSite 0 50 100 150 200 250 Core EdgeNumber of InsectsLocation 0 5 10 15 20 25 30 35 Core EdgeNumber of Isolated SpeciesLocation (d) (a) (b)
Insect Communities in Fragments Laine 6 have low concentrations in the other site locations, except in the case of one Diptera species, which has high abundance in t he core and the edge locations. The four species had even abundances in the continuous forest, while the other sites had differing numbers of each species (Figure 6). Figure 4 The four morphospecies found in every site Figure 5 The distribution of the shared morphospecies across core and edge forest 0 5 10 15 20 25 Core EdgeNumber of InsectsLocation Mogoplistidae Phoridae 1 Phoridae 2 Drosophilidae Mogoplistidae (Scaly Cricket) Phoridae 1 (Humpbacked fly) Phoridae 2 (Humpbacked fly) Drosophilidae (Fruit fly)
Insect Communities in Fragments Laine 7 Figure 6 the distribution of the four shared morphospecies across all sites DISCUSSION Insect communities show ed several trends across the different sites. The number of insects present as well as the number of different morphospecies in each site is relatively constant in every site other than the continuous forest (Figure 1a) (Figure 1b). This serves to clarify the effect that edges and pasture land can have on community compositions as f orest fragments can actually increase the diversity and abundance of insects Fragments usually experience moderate to high levels of disturbance, which have been shown to encourage greater varieties of plants and animals (Ross 2002) (Laurance 2011). While any species loss caused by deforestation is a massive problem the overall communities are seemingly more likely to experience a turnover event as opposed to a net los s, where different species are beginning to occupy niches that have been altered or vacated. Figure 1d shows that the pasture had the most species that were isolated, meaning they were only found in one of the location s This is likely because the pasture offers many habitat characteristics that are not found in forest settings, and thus creates unique morphospecies. This difference b etween forest and pasture is also supported by figure 1c. The fragments and the continuous forest show a similar pattern of insect order composition, likely because they are similar environments, whereas the pasture has a completely different composition of insect orders. While the pasture and the small fragment showed greater diversity than the continuous forest, this does not necessarily mean those communities are stable or healthy. Most forest fragments have only been fragmented recently, and thus have not yet had time t o fully recover from the disturbance and settle into an equilibrium. Davies describes certain traits of populations that lead to a higher probability of extinction in fragments Rare species with low occurrence rates in fragments are more likely to disap pear from the ecosystem in later years than species with high abundances (Davies 2000). The two 0 2 4 6 8 10 12 14 16 Small fragment Large fragment Pasture Continuous forestNumber of InsectsSite Mogoplistidae Phoridae 1 Phoridae 2 Drosophilidae
Insect Communities in Fragments Laine 8 fragments and the pasture had more than double the amount of morphospecies with only one or two individuals sampled (rare species) than the continuous forest di d. This could mean that, while the fragments and pasture technically have higher diversity and abundance at this moment, many species could decline heavily in the future in these locations This pattern, where abundance and richness temporarily increases a nd then rapidly declines, has been shown in previous studies with insects (Larson 2008). The species turnover event, therefore, could only be a temporary phenomenon before species declines. The two forest fragments also exhibited significant differences b etween core and edge locations. Edges had more total insects, and more isola ted species than cores (Figure 3a) (Figure 3 b). Edges are exposed to different conditions than core forest, such as increased sunlight and wind exposure (Murcia 1995) and this cre ates a higher degree o f disturbance and consequently can create higher diversity and abundance (Harper 2005) The forest edge also acts as a meeting point where pasture species and core species can overlap, increasing the overall number of insects sampled. One possible explanation for the high number of isolated species involves total area of habitat for these species. Species that prefer edge habit at are more likely to be sampled than species that prefer core forest habitat, because there is less total edg e area, and therefore a smaller possible range available for the species to be found within. Out of 145 morphospecies, only four were sampled within all study sites. This seems to indicate that each site is separated enough, by distance and environmental conditions, to require differing characteristics among its inhabitance. These differing conditions limit the ability of species to persist across all sites. The four species were also primarily found in core forest conditions, with the greatest concentr ati on in the continuous forest. The four shared morphospecies were also found with relatively even numbers in the continuous forest, while the other sites had inconsistent abundances (Figure 6) This pattern could be an indicator of ecosystem stability. The continuous forest experiences less total disturbance, and has had time to settle into a form of equilibrium. The fragments and pasture have been recently disturbed, from an ecological point of view, and are stil l settling into an equilibrium. Once those l ocations have had more time for the species to find their proper niches and outcompete other species, maybe the pattern of abundance for the fragments and continuous forest will more closely match. The fragments are similar habitats to the continuous fores t, yet the pattern of abundance for the four morphospecies do not match, in the way that the insect order layouts match (Figure 2c). The presence of a greater degree of edge coul d also be skewing these pattern, as the edge habitat creates different species layouts than core (and continuous) forest habitats. It is important to examine why so few species are found across all sites. Three of the shared morphospecies were d iptera, in the families Phoridae and Drosophilidae. The fourth morphospecies is an orthopteran in the family Mogoplistidae. Although not much can be said for the se four specific shared morphospecies, as there can be a lot of variance in the ecology of species within a family, it has been shown that generalist species are better able to proliferate in fragmented landscapes (Larson 2008) At least in the case of Phoridae, they can eat a wide variet y of food sources, and some genera are even parasitoids (Disney 1994).
Insect Communities in Fragments Laine 9 The shared species need to be able to eat many possible food types, an d also need to have a resilient larval stage. Specific requirements for egg laying and larval development could significantly hinder species from proliferating in fragmented sites. This low success rate of cross site species could indicate that most inse cts have some degree of specialization to their environment, and they have not yet had time to adapt to the new, patchy landscape. Future Research Future research should seek to explore the ecological niches that the insects that are found across all fragm ents are filling. Once the environmental role of these insects is characterized, our understanding of the future of these fragments will be enhanced. It is also crucial to repeat this system of sampling across different sites, both in Costa Rica and in other tropical forests. Every fragment and forest has its own context and set of details that could be affecting the data presented here If other fragments show similar patterns, then the shared characteristics of all forest fragments are likely having a gre ater effect than whatever individual characteristics displayed by the location. Lastly, similar research needs to be done with greater taxonomic specificity. Identification to morphospecies helps with comparisons amongst a single collection, but without ac tual species identification the applicability of the research is reduced. Conclusion Forest fragments have far reaching, lasting consequences for all communities of organisms. Fragmentation creates wildly different communities and ecological structures due to the wide variety of habitats that spaw n from deforestation practices. This separates species and restricts their ability to intermix between sites. As humans continue to deforest and fragment, we will shape the evolutionary future of untold numbers of organism s. The species we displace and speciate through the creation of barriers will always adapt to the change Humans, however, may reach the end of their adaptive ability if these deforestation practices continue. It is the hope of this researcher tha t studies similar to this one will build the case against fragmentation, and will lead to greater conservation efforts. But man is a part of nature, and his war against nature is inevitably a war against himself.
Insect Communities in Fragments Laine 10 ACKNOWLEDGMENTS I would like to thank everyone at the Monteverde Institute and the Biological Field Station for providing the means to carry out my research. I would like to thank the Mengel, Arce and Salazar families for allowing me to collect on their lands, and for pre serving small jewels of forest amongst pasture. I would like to thank Emilia Triana and Frank Joyce for their advice and guidance, and for ensuring that I was feeling my oats at all times. I would like to thank Eavy Barbieux for her feedback. I would like to thank Megadeth for providing the soundtrack to my data analysis. Lastly, I would like to thank Aleen Voskanian for being a pillar of support in ways she may never know. LITERATURE CITED Andresen, E. 2003. Effect of Forest Fragmentat ion on Dung Beetle C ommunities and Functional Consequences for Plant Regener ation. Ecography, 26(1), 8797. Davies, K. F., Margules, C. R. and Lawrence, J. F. 2000. Which Traits of Species Predict Population Declines in Experimental Forest Fragments? Ecology, 81: 14501461. D isney R.H.L. 1994. Ecology. In: Scuttle Flies: The Phoridae. Springer, Dordrecht Harper, K. A. et al. 2005. Edge Influence on Forest Structure and Composition in Fragmented Landscapes. Conservation Biology, 19: 768782. Hill, Jane, Michael Gray, Chey Vun K hen, Suzan Benedick, Noel Tawatao, and Keith Hamar. 2011.Ecological impacts of tropical forest fragmentation: how consistent are patterns in spec ies richness and nestedness? Phil. Trans. R. Soc. Hill, J ennifer and Paul Curran. 2003. Area, shape, and isolat ion of tropical forest fragments: effects on tree species diversity and implications for conservation. Journal of Biogeography, Vol. 30, No. 9 pp. 13911403 Jenkins, David et al. 2003. Forest Fragments as Barriers to Fruit Fly Dispersal: Anastrepha (Dipter a: Tephritidae) Populations in Orchards and Adjacent Forest Fragments in Puerto Rico. Environmental Entomology, Vol ume 42, pages 283292 Klein, Bert. 1989. Effects of Forest Fragmentation on Dung and Carrion Beetle Communities in Central A mazonia. Ecology, Vol. 70, No. 6, pp. 17151725 Larsen, Trond H., et al. 2008. Understanding Trait Dependent Community Disassembly: Dung Beetles, Density Funct ions, and Forest Fragmentation. Conservation Biology, vol. 22, no. 5, pp. 12881298.
Insect Communities in Fragments Laine 11 Laurance, William, et al. 2011. The fate of Amazoni an forest fragments: A 32year I nvestigatio n in Biological Conservation. Volume 144, Issue 1, 2011, Pages 5667, ISSN 00063207. Murcia, C. 1995. Edge effects in fragmented forests: implications for conservation. Trends in eco logy & evolution, 10(2), 5862. Ross, K. A. F ox, B. J. and Fox, M. D. 2002. Changes to plant species richness in forest fragments: fragment age, disturbance and fire histo ry may be as important as area. Journa l of Biogeography, 29: 749765. Summerville, K., & Crist, T. 2004. Contra sting effects of habitat quantity and quality on moth communities in fragmented l andscapes. Ecography, 27(1), 312.
Insect Communities in Fragments Laine 12