1 Evaluating Windbreaks as a Conservation Strategy for Pollinator Communities in Altered and Fragmented Landscapes Anna Gouznova Department of Environmental Studies & Department of Political Science, Fordham University, 113 W 60 th Street, New York, New York 10023, USA ABSTRACT In Costa Rica, sustaining human population growth has required fragmentation and conversion of its Tropical landscape to an agricultural one. Such large scale land use changes decrease local abundance of species and disrupt im portant ecological relationships, including species richness and productivity within pollinator communities (Kearns et al. 1997). Pollinators are essential in maintaining stability of plant communities and their dependent consumers. Moreover, in specialize d pollinator host plant mutualisms, the loss of either will reduce or eradicate population of the other. To counter effects of habitat fragmentation, agricultural communities incorporated windbreaks into crop fields and pastures as a restorative effort. I evaluate the performance of windbreaks as repository for pollinators in the San Luis Valley of Puntarenas, Costa Rica by assessing richness and abundance of both pollinator and insect communities within three distinct windbreaks and a bordering secondary g rowth forest. Pollinator species richness was highest in the forest (H`= 3.72), along with superior numbers of families and individuals. Two species of non insect pollinators, hummingbirds Amazilla tzacatl and Hylocharis eliciae were never seen in the fo rest but regularly in two of the three windbreaks. The two sites, one in a shade grown, Coffea arabica (Rubiaceae) farm and the second, in a pasture, supported comparable species richness (H` = 2.65 and H` = 2.93), and while a slightly lower amount of indi viduals was collected in the farm site, it contained nine more families than the pasture. In the same pasture, the third windbreak exhibited lowest diversity (H` = 2.65), and nearly half the number of families and individuals found in the first, neighborin g pasture site. Higher pollinator richness and abundance correlated with lower mean temperatures of about 26C and wind speeds in the range of 0.38 m/s 0.9 m/s. Overall, structural attributes, including plant species composition, presence of bodies of wa ter and abiotic conditions, determined a windbreak's capacity to retain pollinators in an altered landscape. RESUMEN En Costa Rica, el crecimiento sostenible de la poblacion humana requirÂŽ la fragmentaciÂ—n de los bosques y la conversiÂ—n a campos agrico los. Tales cambios en la utilizaciÂ—n del suelo a una escala grande disminuyen la abundancia local de especieje
2 interrumpen alguna relaciones ecolÂ—gicas importantes, incluyendo riqueza de la especie y la productividad dentro de comunidades del polinizadores (Kearns et el al. 1997). Los polinizadors son esenciales en la estabilidad que mantienen de las comunidades de las planta y de sus consumidores. Por otra parte, en mutualisms especializados reducÂŽ o suprimÂŽ la poblacion de la otra. Para contradecir los ef ectos de la fragmentaciÂ—n del habitat, las comunidades agracolas han incorporadÂ— cortavientos en campos de la cosecha y pastos como esfuerzo restaurativo. EvaluÂŽ el funcionamiento de cortavientos como deposito para los polinizadores en San Luis, Puntarenas Costa Rica. DeterminandÂŽ riqueza y abundancia de comunidades del polinizadores y del insectos dentro de tres cortavientos distintos y de un bosque secundario del crecimiento que confina. La riqueza de la especie del pollinator era la mÂ‡s alta del bosque, H` = 3.72, junto con numeros superiores de familias y de individuos. Sin embargo, dos especies de aves, Amazilla tzacatl y Hylocharis eliciae nunca fueron consideradas en el bosque pero regularmente en dos de los tres windbreaks. De este en una cortina c recida, la granja del arabica de Coffea y la otra en un pasto, apoyaron la riqueza comparable de la especie (H` = 2.65 y H` = 2.93). Mientras que recogiÂŽron levemente a menos individuos en el sitio de la granja, contuvo a nueve mas familias que el pasto. E n el mismo pasto, el tercer windbreak exhibite la diversidad mas baja, H` = 2.65, y casi la mitad del numero de familias y de individuos encontre en el primer, sitio vecino del pasto. H. eliciae fueron observados a una vez. Una riqueza y una abundancia ma s altas del pollinator correlacionaron con temperaturas malas alrededor de 26C y velocidades del viento con en la gama de 0.38 m/s 0.9 m/s. Una abundancia mÂ‡s alta de insectos correlacionÂ— linear con la abundancÂ’a de pollinators totales, cualidades estr ucturales, incluyendÂ— la composiciÂ—n de la especie de la planta, presencia de aguas de superficie y las condiciones abiotic, determinaron la capacidad de un cortaviento de conservar polinizatores en un paisaje alterado. INTRODUCTION Frafmentation and co nversion of forest landscapes to agriculture decreases availability of nesting sites and flower food sources for native pollinators, consequentially decreasing their richness, abundance and productivity (Kearns et al. 1997; Mustajarvi et al. 2001). As a re sult, dependent plant communities experience reduced seed sets, production of less fit offspring and ultimately, extinction (Didham et al. 1996). Evolution of the unique interaction between pollinators and host plants contributes to its value. At least 93% of subcanopy and understorey plant species depend on insects for pollination, and to these, "pollinators are as critical as light and water" (Bawa, 1990; Levin 1971). Evolved morphologies, for example the sucking proboscises of Lepidoptera, limi t the range of host plants available for a pollinator species (Barth 1991). Floral specialization limits pollinator access to flowers through complex shapes, size, corolla, and floral design (Faegri and van der Pijl 1979; Linhart and Feinsinger 1980) or sp ecialized attractants and rewards (Dafni et al. 1990; Johnson and Bon 1993; Steiner 1989). Unnatural changes in both size and spatial arrangement of plant populations, can also define demographics and genetics of a pollinator community, with plan t species
3 density having a greater effect than area (Mustajarvi et al. 2001). Smaller, fragmented plant communities regularly suffer from insufficient pollen transfer and consequently lower seed sets as well as higher levels of inbreeding (Kearns et al. 19 97). These aftereffects may be averted by planting windbreaks, which amplify conservation efforts by serving as habitats for many species of forest trees (Harvey 2001). More than half a million trees have been planted in windbreaks by the Monteverde Conser vation League, including those studied in San Luis (Burlingame 2001). Studied windbreaks varied in area, structure and plant species composition. Sites in the pasture, islated from both adjacent forest and farms, were contrasted with windbreaks in Coffea arabica plots, planted among several species of fruit and shade trees. When combined with other attributes, shade grown plots parallel to structural designs of forests and higher levels of biodiversity in birds and arthropods (Perfecto et al. 1996) Here I expected higher richness and abundance of pollinators. Areas of higher plant density are correlated with higher abundances of most insects and hummingbirds, and accordingly, pollinators. In pastures, fast growing windbreaks are more desirable and are usually limited to two or three species such as cypress, Cupressus lusitanical (Cupressaceae), Tubu, Montanoa dumicol, (Asteraceae) and pine, Pinus spp (Pinaceae). My research evaluates which factors determine the success of windbreaks as repositories o f pollinators, including effects of abiotic conditions and plant species richness on abundance and richness pollinators, as well as any correlations between insect and pollinator communities. METHODS Study Sites Research was carried out in th e San Luis Valley of Puntarenas, Costa Rica between April 14 th and May 5 th of 2007. Located below the continental divide on the leeward side of the Tilaran Range, San Luis Valley is approximately 1,200 m above sea level in the Pre Montane Wet Forest, accor ding to Holdridge Life Zones (Bolanos et al. 1993). Four sites were selected, of which three represented different windbreak structures and a site in secondary growth forest was thought to harbor most remaining insects and pollinators. Pasture P ine: One of two windbreaks located in Rafael Leiton's pasture. This windbreak is designated as Pasture Pine, referring to dominance of Pinus spp. (Pinaceae), with only eight other plant species present. Study site in Pasture Pine was 146 m long 8.5 m wide and entirely intersected the pasture. Two watering canals for cattle, each about 2 m wide, run through this site, one directly behind and slightly intertwined with planted Pinus spp. the other spans the widths of study site and pasture (See Fig. 1) Pasture Fruit: The second site is also located in Rafael Leiton's pasture. Chiefly composed of common, neotropical fruit trees, it was denoted as Pasture Fruit, and contained a total of 11 observed species. This windbreak runs parallel to Pasture Pine a nd while, spans the entire length of the pasture it becomes too sparse, restricting actual study area to 64 m in length and 5.5 m in width (See Fig 1). The third site is located in Finca La Bella, as part of a plot owned and managed by Gilberth L obo, who cultivates Coffea arabica (Rubiaceae). It is 71 m long and 5.2 m
4 wide. This windbreak has a stream running along its edge. Planted on both sides of C. Arabica are assorted banana palms, Musa acuminata (Musaceae) and several varieties of lemons, C itrus aurantium (Rutaceae) and Citrus x paradisi (Rutaceae) as a natural source of shade, for a total of 19 plant species. Accordingly, this windbreak is designated as Shade Grown Farm. The fourth site was set up approximately 37 m into a second ary forest adjacent to another pasture, directly across the road from Rafael Leiton's land. In relation to the windbreaks, study area here was smallest, 11 m by 12 m, yet hosts at least 26 plant species within the perimeter. This area includes a small, wel l regenerated light gap. Area was limited due to high barbed wire along entry points, uncompromising density of vegetation and branches out of reach where I would typically arrange my traps. FIGURE 1. Distribution of study areas in San Luis Valley, Pu ntarenas, Costa Rica. Pasture Pine (A) and Pasture Fruit (B) run parallel, about 40 m apart. Area studied in Forest is reflected by D. Shade Grown Farm (C) spans the width of Gilberth Lobo's plot. Materials Five traps were set up in each site: t wo Standard Bottle Hummingbird Feeders, one PIN Bee Trap, and two Aeroxon Flycatchers, flytraps. Hummingbird feeders were filled with a two part sugar to five part water solution. Half cup honey and two tablespoons of canned Tuna fish were placed in each b ee trap, inner and outer walls were sprayed with a pulverizer containing a one part honey to five parts water solution. Flytraps were affixed to branches with strips of string. Locations of traps were reselected after each collection, so that each site was sampled in its entirety. Hummingbird feeders and bee traps were placed approximately three meters apart, with one flytrap accompanying a bee trap to
5 maximize visitor capture. The second flytrap was positioned between the hummingbird feeder and bee trap. Methods Every six days, all specimens were extracted from all traps for a total of three collections. Hummingbird feeder frequently contained ants (Formicidae), that were also collected. Specimens from flytraps were more susceptible to damage an d extracted with special care. Individuals from bee traps were repeatedly flushed with a strong detergent to remove oil film from Tuna. Specimens were transferred to labeled vials with alcohol and stored in freezer until future use. In between collections, field observations were made for a total of seven observation days. During observation days, in the beginning of each transect, temperature was recorded using a non digital, mercury thermometer and wind speeds were recorded with Kestrel 3000, on the leewa rd sides. I spent 30 minutes in each site, surveying the entire area, recording any pollinators or insects, especially hummingbirds which could not have been quantified otherwise. Additionally, I attempted to catch as many insects as possible with a hand h eld, one m handle long net. Data from observations were merged with collections records and all specimens were identified to order, family and species/morpho species using an Olympus dissecting microscope and Charles L. Hogue's Latin American Insects and E ntomology Jim Wolfe, an entomologist, reviewed and supplemented my lists. RESULTS Influence of Ambient Conditions Ambient temperatures and wind speeds varied significantly despite proximity between sites (Figs. 2 & 3). Mean temperatures were highest in Pasture Fruit (29.1C), followed by Pasture Pine (26.8C), second lowest were recorded in Shade Grown Farm (26.5C) and lowest mean temperatures were noted in Forest (26C). Sites in the pasture, which experienced a substantial difference in both mean temp eratures and wind speeds, are only 40 m apart, signaling that something inherent to the windbreak must be responsible for lower temperatures in Pasture Pine. Reasonably, highest wind speeds (1.96 m/s 4.7 m/s) noted in Pasture Pine, reduced ambient temper atures, along with shade from its tall Pinus spp (See Fig. 3) However, high wind speeds were not compulsory for lower mean temperatures. Daily averages in Pasture Pine were still second hi ghest (26.8C), and without wind gusts would have mirrored temperatures in Pasture Fruit. Actually, sites with lowest mean temperatures, Shade Grown Farm (26.5C) and Forest (26C), were most dense in their plant composition. Accordingly, less dense windbr eaks in the pasture experienced highest mean values.
6 24 25 26 27 28 29 30 Pasture-Pine Pasture-Fruit Shade GrownFarm Forest Temperature Â¡C FIGURE 2. Mean temperatures recorded in all sites. Highest mean temperatures were recorded in Pasture Fruit (29.1C), followed by Pasture Pine (26.8C). Second lowest mean was recorded in Shade Grown Farm (26.5C).Lowest mean temperatures were recorded in Forest (26C). 0 1 2 3 4 5 6 7 Pasture-Pine Pasture-Fruit Shade GrownFarm Forest Wind Speed m/s Min. Max. FIGURE 3. Mean maximum and minimum wind speeds in all sites. Highest mean wind speeds were recorded in Pasture Pine (1.96 m/s 4.7 m/s). Pasture Fruit is second highest (0.62 m/s 1.98 m/s), followed by Shade Grown Farm (0.36 0.6 m/s). Mean wind speeds are lowest in the forest (0.38 m/s 0.9 m/s).
7 Pollinator and Insect Demographics Highest numbers of families, species and individuals for both insects and pollinators were found in Forest. Out of 176 individuals collected, 73 represented 51 species of pollinators, more than double the number of pollinating species in all other sites (See Figs. 4 & 5.) Thirty two families of pollinators were identified ou t of 56 insect families collected Chi square values found abundance of insects to be significant across all three categories, individual specimens ( x2 = 55.693 df = 3, P <0.05), species (x2 = 31.8495, df = 3, P <0.05) and family (x2 = 14.6685, df = 3, P <0.05). In relation to pollinators, Chi square values were significant for number of individuals (73), (x2 = 20.0057, df = 3, P <0.05), as well as family composition (x2 = 11.44, df = 3, P <0.05). Abundance of insects and pollinators was similar between Pasture Pine and Shade Grown Farm, so I contrast both sites. More insects were found in Shade Grown Farm than in Pasture Fruit. Out of 104 individual insects found in Shade Grown Farm, 37 were pollinators, while out of 83 insects in Pasture Fruit, 44 were pollinators. Nineteen families of insects were found in Pasture Fruit, less than half of the 42 families found in Shade Grown Farm. This accounted for several (3) more species of insects and pollinators (2) found in Shade Grown Farm. In both sites, pollinator abundance was represented by similar numbers of families, species and individuals. However, Pasture Fruit contained seven more pollinator families that Shade Grown Farm. Overall, abundance of both insects and pollinators was comparable between Pasture Fruit and Shade Grown Farm. Nearby, Pasture Pine supported less than half the number of species (19) than its neighboring site, Pasture Fruit (46). Interestingly, only five more pollinator species were found in Pasture Fruit, with an ana logous number of families. Number of insects (41) and pollinators (20) collected was less than half of individual insects (83) and pollinators (44) found in the Pasture Fruit. Less than one fourth the nunmber of insects found in Forest (176) was found in P asture Pine (41). Four times as many pollinator families were found in Forest. Pasture Pine had lowest indicators for both insects and pollinators across all categories. This explains why significant Chi square values were found for the number of individua ls belonging to the general insect community, ( X 2 = 35.6435 df = 3, P <0.05), and the number of species, ( X 2 = 20.5097, df = 3, P <0.05). Chi square value was also significant for the number of pollinators present, ( X 2 = 12.6954, df= 3, P <0.05).
8 18 19 42 56 19 46 49 92 41 83 104 176 0 20 40 60 80 100 120 140 160 180 200 Pasture-Pine Pasture-Fruit Shade GrownFarm Forest Number of Individuals Families Species Individuals ________________________________________________ FIGURE 4. Distribution of pollinators collected and observed across all sites. Pollinators were most abundant in Forest. In windbreaks, similar number of species were recorded in Pasture Fruit (21) and Shade Grown Farm (23). Although the number of families (19) and individuals (44) was greater in Pasture Fruit than Shade Grown Farm, similar values were found for species. Pasture Pine was least diverse, comprised of 8 families, 1 6 species and 20 individuals. 18 19 42 56 19 46 49 92 41 83 104 176 0 20 40 60 80 100 120 140 160 180 200 Pasture-Pine Pasture-Fruit Shade GrownFarm Forest Number of Individuals Families Species Individuals FIGURE 5. Distribution of insects collected and observed across all sites. Most individuals (176), species (92) and families (56) were found in the Forest. Pasture Pine was the least abundant s ite across all three categories with only 41 individuals, 19 species and 18 families. More individuals (104) and families (42) were collected in Shade Grown Farm than in Pasture Fruit, but only three more species were present in Shade Grown Farm.
9 Pollinators Hylocharis eliciae and Amazilla tzacatl hummingbirds, followed distinct trends (See Fig.6). Surprisingly, neither species was seen in the Forest, where highest abundance and diversity of both insects and pollinators were recorded. Both species were observed everyday in Shade Grown Farm and occasionally in Pasture Fruit. This can be attributed to higher density of fruiting and flowering trees in both sites. While, Hylocharis eliciae was seen in Pasture Pine, I believe it mi grated from Pasture Fruit, since no other sightings occurred. 0 3 11 0 1 4 8 0 0 2 4 6 8 10 12 Pasture-Pine Pasture-Fruit Shade GrownFarm Forest Number of Individuals Amazilla tzacatl Hylocharis eliciae FIGURE 6. Non insect pollinators: sightings of hummingbirds, Amazilla tzacatl and Hylocharis eliciae across all sites. Both species were frequently observed in Sha de Grown Farm while none were seen in Forest. Both species were also recorded in Pasture Fruit, from which one individual, H. eliciae likely migrated to Pasture Pine at time of observation.
10 Overall, species richness was hig hest in Forest site, (H` = 3.72), lowest in Pasture Pine, (H` = 2.57), second lowest is Pasture Fruit, (H` = 2.65,) and second highest in Shade Grown Farm, (H` = 2.93) (See Fig. 7). 2.57 2.65 2.93 3.72 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 Pasture-Pine Pasture-Fruit Shade GrownFarm Forest H` Value FIGURE 7. H` values for pollinator richness in all sites. From lowest to highest: Pasture Pine (H` = 2.57); Pasture Fruit (H` = 2.65); Shade Grown Farm, (H` = 2.93); and Forest (H` = 3.72). Pollinator and insect communities were distinct within each site. When Modified t test values were calculated for all p airwise combinations of study sites, almost all were significant. Statistical differences were present between Pasture Pine and Shade Grown Farm sites, (Modified t test = 2.19, df = 54.69, P <0.05); Pasture Pine and Forest sites, (Modified t test = 8.96, df = 59.05, P <0.05); Pasture Fruit and Shade Grown Farm sites, (Modified t test = 1.93, df = 73.83, P <0.05); Pasture Fruit and Forest, (Modified t test = 6.02, df = 83.21, P <0.05). No significant difference was found between Pasture Pine and Pasture Fru it sites, (Modified t test =.42, df = 62.44).
11 Correlations Between Insects and Pollinators Moreover, abundance of families, species and individuals of either pollinators or insects correlated with one another. This is a meaningful f actor in conservation biology for pollinator communities. Abundance of pollinators corresponded to increases in insect populations. Even Pasture Pine, with only 41 total insect individuals collected this trend persists. In all sites, a linear correlation c an be observed between abundance of both pollinators and insects, regardless of sample size (See Fig.8). 0 10 20 30 40 50 60 70 80 90 Families Species Individuals Pasture-Fruit Number of Individuals 0 5 10 15 20 25 30 35 40 45 Families Species Individuals Pasture-Pine Number of Individuals Insect Pollinator 0 20 40 60 80 100 120 Families Species Individuals Shade Grown-Farm Number of Individuals 0 50 100 150 200 Families Species Individuals Forest Number of Individuals FIGURE 8. Correlations between richness and abundance of insects and pollinators in each site. Higher numbers of families, species and individuals found for insect communities correlated with higher numbers for pollinators across all three categories. This trend was observed regardless of sample size. Breakdown of Insect Orders and Families In Pasture Pine, Diptera was the main orde r found, but since it consists of flies, had only one potential pollinator family, Cauliphoridae (See Fig. 9). Blattaria, another large order, consisted entire of one family of cockroaches, Blattidae. Lepidoptera found here were mostly comprised of Rioniid ae, and were usually seen only next to the fruits of a random Psidium guajava (Myrtaceae) in this site.
12 Orders Represented in Pasture-Pine Blattaria Coleoptera Diptera Odonata Apodiformes Lepidoptera Families Represented in Pasture-Pine Blattidae Cauliphoridae Coenagrionidae Culicidae Other Formicidae Lauxaniidae Lycaenidae Microphysidae Muscidae Nymphalidae Unknown Pieridae Rioniidae Tephritidae Traigonidae Vespidae Trachelidae FIGURE 9. Orders and families of insects represented in Pasture Pine. Diptera, order of flies, was the most prevalent order in Pasture Pine. Famil ies of this order included Cauliphoridae, Lauxaniidae, Microphysidae, Muscidae, Tephritidae and Trachelidae. Blattaria, order of cockroaches was the second most abundant, followed by Lepidoptera, moths and butterflies. Orders such as Hymenoptera and Lepidoptera were most prominent, followed by Blattaria. In Hymenoptera, members of Vespidae are great pollinators. More families of Lepidoptera were found in this site than in Pasture Pine, including, Nymphalidae, Rioniidae and Pieridae. Coloptera was also a significant order with many pollinating beetle families such as Carabidae, Chrysomelidae, Erotylidae and Nitidulidae (See Fig. 10). Orders Represented in Pasture-Fruit Apediformes Blattaria Coleoptera Damselfy Diptera Homoptera Hymenoptera Lepidoptera Odonata Orthoptera Families Represented in Pasture-Fruit Blattidea Calliphoridae Campthoridae Carabidae Chrysomelidae Coenagrionidae Drosophilidae Erotylidae Formicidae Lauxanidae Lycaenidae Microphysidae Muscidae Neriidae Nitidulidae Nymphalidae Unknnow Pieridae Rioniidae Tracheliade Vespidae FIGURE 10. Orders and families of insects represented in Pasture Fruit. Diptera comprised majority of insects found in Pasture Fruit. Hymenoptera, inclusive of ants, bees and wasps was second most abundant. Lepidoptera was also significantly represented. A large number of families was represented in this site.
13 Orders in Shade Grown Farm includ ed Lepidoptera, Hymenoptera and Diptera. Trachelidae, a family of flies, was well represented in this windbreak, but is not a pollinator. Other non pollinating fly families were also common, including Calliphoridae (See Fig. 11). Orders Represented in Shade Grown-Farm Anoplura Blattaria Coleoptera Diptera Apediformes Hymenoptera Lepidoptera Families Represented in Shade Grown-Farm Blattidae Calliphoridae Carabidae Cercopidae Chrysomelidae Coccinellidae Ctenidae Elateroidea Formicidae Gryllidae Ithomiinae Lampyridae Lycaenidae Trachelidae Vespidae FIGURE 11. Orders a nd families of insects represented in Shade Grown Farm. Diptera was the main order found, followed by Lepidoptera and Hymenoptera. Diptera families found include Calliphoridae and Trachedidae. In Lepidoptera, Lycaenidae and Ithomiinae were main families fo und. Hymenoptera were represented by Formicidae and Vespidae. Most orders were seen in Forest, as well as the largest number of families. Lepidoptera represented nearly half of insects collected and observed, and are very good pollinators. Lep idoptera families not found in other sites include Saturnidae. Coleoptera comprised a large portion of insects, with numerous families unique to the site such as Cydnidae and Curculionidae. Many families of pollinators were evenly represented in Forest (Se e Fig. 12).
14 Orders Represented in Forest Blattaria Coleoptera Dermoptera Diptera Hemiptera Hymenoptera Lepidoptera Micro Lepidoptera Odonata Orthoptera Unknown Families Represented in Forest Asciidae Battidae Blaberidae Calliphoridae Carabidae Cecropidae Chrysomelidae Coenagrionidae Nymphalidae Curculionidae Cydnidae Ellatirodae Gryllidae Halictidae Ichnomonidae Irophoridae Lampyridae Languridae Lepturinae Lygaeidae Membrasidae Microphysidae Muscidae Neriidae Nitidulidae Nymphalidae Psychedelidae Pyrrhocoridae Rioniidae Saturnidae Scalitidae Scarabidae Syrphyd Tabanidae Techenidae Therevidae Unknown Vespidae FIGURE 12. Orders and families of insects represented in Forest. Lepidoptera were most common in Forest, along with Coleoptera, Diptera and Hemiptera. This site represented over 56 families. DICUSSION Conversion and fragmentation of natural landscapes carries important implications for ecosystem functions (Didham et al. 1996). In isolated patches, pollinators find themselves in "ecological traps." Once the number of pollinators has been reduced, remaining plants suffer from reprodu ctive failure, degrading the plant community even further (Jennersten et al. 1998). Additionally, flowers in small populations either receive fewer visits from pollinators or receive pollen from sibling plants, yielding low seed production (Kearns et al. 1 997). Since individual species of pollinators differ in their responses to habitat loss, a wide and unpredictable array of potential effects can be observed at the community level (Robinson et al. 1992). Overall, management activities that threaten pollina tor populations must be avoided in order to protect populations of endangered plants (Daily). Undeniably, declines within pollinator communities are not restricted to insect orders and reverberate throughout an ecosystem.
15 When areas of the Guanac aste Province of Costa Rica were deforested for cattle grazing, dramatic reductions in abundance and richness of bees followed (Vinson et al. 1993). Powell and Powell elaborate that "a mosaic of forest fragments surrounded by treeless pastures will not sup port populations of some bee species," referring to 15 species of Euglossine bees co evolved with many species of Orchidaceae (Powell and Powell 1987). In their study, isolated patches experienced declines in visitation rates of males for 15 bee species. I n an agricultural community such as San Luis, these trends may be widespread. Extensive land conversion, as in Guanacaste, requires conservation efforts to avert long term environmental pressures on pollinator communities. Patterns of diversity i n pollinator communities were unique to each site and responded to multiple abiotic factors as well as enrichment from the greater insect community. As the number of families, species and individuals of insects increased, pollinator communities experienced higher values in corresponding categories. This linear relationship supports conservation of the entire insect community, not just individual species of pollinators. I evaluate how both insects and pollinators Several observed trends are particularly enga ging. Despite their dissimilarity in structure and location, abundance of pollinators in Pasture Fruit and Shade Grown Farm was rather congruent. Two additional species contributed to slightly higher species richness in Shade Grown Farm, but seven more fam ilies and individuals were found in Pasture Fruit. Conversely, Pasture Pine, situated only about 40 m away from Pasture Fruit, supported half the number of pollinating individuals of its neighbor, as well as a lower number of species and families. Distinct variations suggest that traits pertinent to sites influence their capacity to sustain pollinator communities. Structural advantages can be seen in Shade Grown Farm, situated in Finca La Bella. Finca La Bella is a major source of harvested Co ffea arabica for CafÂŽ Monteverde. These growers use little to no agrochemicals and integrate a variety of crops into their plots. Farmers have planted or maintained existing natural windbreaks to protect cattle from detrimental effects of strong winds, inc luding soil erosion and as sources of shade for their crops. Gilberth communicated to me that many farmers acknowledged higher plant diversity and density to attract more pollinators than monocrop fields, increasing reproductive success and stability of th eir primary crop, C. arabica. Gilberth's plot is dense with a high number of plant species and has a small stream running along the windbreak, a potential breeding ground for both insects and pollinators. My study period coincided with synchronous blooming of Coffea in April and May, dormant during the dry season. Coffea phenology might have drawn more pollinators to this windbreak, but their presence is best explained by structural features of Shade Grown Farm. Gilberth does not immediately clear fallen M usa acuminata brackets or snapped twigs, which frequently serve as nesting sites. Fallen fruit from Citrus aurantium and C. x paradisi is left to naturally decompose. Smaller Lepidoptera can be seen floating around the masses of fruit. A higher concentrati on of fallen fruit and flowers from Psidium guajava (Myrtaceae) and C. aurantium (Rutaceae) was likely responsible for attracting pollinators to Pasture Fruit. In contrast to Finca La Bella, context for Pasture Pine perpetuated lower pollinator d iversity. Pasture Pine, had several slow running, streams of potable water nearby, comparable to Shade Grown Farm. Even with the possibility of migrants from Pasture Fruit, this windbreak exhibited the lowest number of families, species and individuals of
16 both insects and pollinators. Pasture Pine appears to lack capacity to retain a diverse set of pollinators because it is dominated by Pinus spp (Pinaceae). Pollinator richness and abundance were notably highest in the forest which was conspicuously denser and hosted at least three times as many plant species as Pasture Pine. From many structural attributes of windbreaks, data consistently highlight plant species richness as an indicator of a healthy pollinator community. Abiotic conditions are als o precursors for performance of windbreaks as a conservation effort. Pasture windbreaks experienced highest minimum and maximum wind speeds and highest mean temperatures (Fig. 2 and Fig. 3). As mentioned earlier, both are in rather close proximity, but hot test temperatures, among all sites in fact, were recorded in Pasture Fruit, on average, 2.3Â¡C higher. Pasture Pine is likely experiencing lower temperatures due to shade cover and higher wind speeds. However, hotter ambient temperatures did not discourage pollinator visits in Pasture Fruit nor did lower temperatures encourage diversity in Pasture Pine. During field observations, Lepidoptera did not approach Pasture Pine after visiting Pasture Fruit, and fewer species were recorded overall. From observations I attribute this behavior to powerful winds in Pasture Pine. Coleopterans, only one species was found in this windbreak, may have been disproportionately affected by such variations in air temperatures and litter moisture content, also dependent upon amb ient temperatures. Species richness of Coleoptera has been found to be least similar between edge sites and continuous forest (Didham et al. 1996). Generally, lower wind speeds and temperatures of about 26Â¡C appear to be optimal for both pollinators and in sects and should be propagated by increasing density within existing windbreaks. Pollination is an ecological service that relies on diverse ecological communities for its success (Daly et al.1998). However, even in a heavily altered and used landscape, moderately sized patches of high resource quality may be equivalent to larger patches of lower quality (Metzger and Decamps 1997.) As in the case of Shade Grown Farm and Pasture sites, the windbreak area in C. arabica plots is smaller but s upports higher diversity and abundance within both pollinator and insect communities. Thus windbreaks can be implemented successfully as conservation strategies if properly designed. Deforested areas, such as pastures, between remnant forest have long been know to impose barriers that discourage movement of many forest dwelling species. Nielsen found that windbreaks successfully serve as corridors for many birds species. However, these windbreaks were composed of only two tree species, cypress, Cu pressus lusitanical (Cupressaceae) and Tubu, Montanoa guatemalensis (Asteraceae), neither of which produced fruit, nectar or seeds at time the study was conducted, yet insect eating birds comprised almost one half of the species captured (Nielsen et al. 20 01). Clearly these birds must have been foraging on insects present in the windbreaks. In striving to enrich pollinator and insects communities by establishing windbreaks, other species will also benefit. In 1975, pollination was considered under studied, a "weak link in our understanding of how ecological communities function (Kevan 1975)". Today, a wide range of available literature allowed me to clarify trends in diversity within pollinator communities in a modified landscape. So far, my definit ion of the nature and value of this ecosystem service has been expressed primarily through its disruption and loss, emphasizing the need to reassess windbreaks as repositories for pollinators (Daily 1997).
17 My evaluation suggests that many windbreaks curren tly lack structural attributes, namely adequate host plant communities, and will not be successful in maintaining diversity and abundance of pollinators. I recommend that existing windbreaks be substantially filled with more and native plant species, not m embers of Pinus. By increasing plant density and diversity, shade cover will reduce ambient temperatures and wind speeds, while new rewards offered will attract more pollinators. Windbreaks should connect to forest or remnants, thereby, creating migratory corridors and enhancing genetic variability of pollinators, as well that of their hosts. My statistical assessment of pollinator communities in fragmented and significantly altered habitats corroborates with my proposal and should serve to improve the outc ome of windbreaks as a conservation strategy. Ideally, further studies would inquire into structural dynamics that best promote diversification of pollinator and insect demographics. A database needs to be created of insect and pollinator species found in nearby forest, less than 100 m inland from the edge, and should be cross listed with species found in windbreaks. Host plants of species absent from windbreaks in proximity to the forest should also be identified. These plant species should be ad ded to already existing windbreaks and monitored. Linking habitats with windbreaks may help maintain biodiversity by enabling use of forest fragments that otherwise would have been left as isolated remnants (Nielse et al. 2001).Concerned residents and othe r community members should be informed and invited to participate in the restoration process with assistance from experienced organizations such as the Monteverde Conservation League. ACKNOWLEDGMETS Dr. Alan Masters, first and foremost I thank you for a lways expressing confidence in me and in all of us. I am obliged to Dr. Karen Masters for leading me through initial stages of project anarchy. Of course, this study would not have been completed without Jim Wolfe's expertise and generous assistance. For t heir 24 hour support, I revere Camryn Pennington and Thomas McFarland. I express my gratitude to Rafael Leiton for granting access to his pastures and improving my cow milking technique. Additionally, I thank Hilbert for use of his plot in Finca La Bella. I applaud Dr. John Davenport for his uncompromising guidance. And for always being there while rarely on the same continent, I wish to dedicate my work to Salvatore J. Grasso. LITERATURE CITED Barth, Friedrich G. 1991. Insects and Flowers: The Biology o f a Partnership. Princeton University Press, Princeton, New Jersey. Bolanos, R. A., and V.Watson. 1993. Mapa ecologico de Costa Rica. Centro Cientifico Tropical, San Jose, Costa Rica. Burlingame, Leslie J. 2001. Conservation in th e Monteverde Zone: Contributions of Conservation Organizations. In: Monteverde: Ecology and Coservation of a Tropical Cloud Forest Nadkarni M. Nalini, eds. Oxford University Press, Oxford, UK pp. 365 366 Carriere, Jean. 1991. E nvironment and Development in Latin America: The Politics of Sustainability. Manchester: Manchester University Press, page 6. Dafni, A.1992. Pollination ecology: a practical approach. Journal of Ecology, 83, 2 : 347 348. Oxford University Pr ess, Oxford, UK. Daily, Gretchen C. 1997. Nature's Services Island Press, Washington, D.C., Island Press, pp 5, 142 145.
18 Daly, Howell, John T. Purcell and H. Alexandra. 1998. Introduction to Insect Biology and Diversity. Oxford University Pre ss, Oxford, UK, page 28. Didham, Raphae K; Peter M. Hammond, John H.Hawton, Paul Eggleton and Nigel E. Stork. 1998. Ecological Monographs, Vol. 68, 3 :295 323. Faegri, K and van der Pijl, L. 1979. The principles of pollination ecology. Oxfor d: Pergamon press. Harvey, Celia. 1999. Windbreaks as Habitats for Trees. In: Monteverde: Ecology and Coservation of a Tropical Cloud Forest Nadkarni, M Nalini, eds. Oxford University Press, Oxford, UK, pp. 450 451. Hogue, Ch arles L. 1993. Latin American Insects and Entomology England, Oxford, University of California Press. Jennersten O., J. Loman, A. P. Moller and B. Widen. 1997. Conservation biology in agricultural habitat islands. Eco. Bull 46 :7 2 87 Johnson, S.D. 1992. Plant animal relationships. In: The Ecology of fynbos. R. M. Cowling, ed. Oxford University Press, Oxford, UK, pp. 175 205. Kearns, Carol Ann and Inouye, David William.1997. Pollinators, Flowering Plants and Conserv ation Biology BioScience 47,5: 297 307 Kevan, P.G.1975. Pollination and environmental conservation. Environmental Conservation 4: 293 298. Linhart, Y.B. and Feinsinger, P. 1980. Plant hummingbird interactions: effects of island size and degree of specialization on pollination. Journal of Ecology 68 :745 760. Metzger, J.P., and H. Decamps. 1997. The Structural Connectivity Threshold: A Hypothesis in Conservation Biology at the Landscape Scale. Acta Oecologica International Journal of Ecology 18 :1 12. Mustajarvi, Kaisa and Pirkko Siikamaki, Saara Rytkopmen, and Antti Lammi, 2001. Consequences of plant population size and density for plant pollinator interactions and plant performanc e. Journal of Ecology : 80 87, 89. Nielsen, Karen and Debra DeRosier. Windbreaks as Corridors for Birds. In: Monteverde: Ecology and Coservation of a Tropical Cloud Forest Nadkarni, M Nalini, eds. Oxford University Press, Oxford, UK, pp. 448 450. Perfecto, Ivette, Robert A Rice, Russel Greenberg and Martha van der Voort.1996. Shade Coffee: A Disappearing Refuge for Biodiversity. Bio Science, Vol 46 8: 598 608. Perfecto, Ivette and Roy Snelling. 199 5. Biodiversity and the Transformation of a Tropical Agroecosystem: Ants in Coffee Plantantions. Ecological Applications 5 :1084 1097. Perfecto, Ivette, John Vandermeer, Alex Pinto Mas and Lorena Soto.2005. Biodiversity, yield and shade coffee certification. Ecological Economics 54 : 435 446. Powell, Harriet A. and George V.N. Powell. 1987. Population Dynamics of Male Euglossine Bees in Amazonian Forest Fragments. Biotropica, Vol. 19, 2 :176 179. Proctor, M., P. Y eo and A. Lack. 1996. In: Monteverde: Ecology and Coservation of a Tropical Cloud Forest Nadkarni, M Nalini, eds. Oxford University Press, New York, New YorkOxford, pp 246 247. Robinson, George R, Robert D Holt, Michael S. Ga ines, Steven P. Hamburg, Michael L. Johnson, Henry S. Fitch and Edward A. Martinko, Edward A. 1992. Diverse and Contrasting Effects of Habitat Fragmentation. Science Vol 257, no 5069: 524 526. Steiner, K. E. 1989. Pollina tion of Disperis (Orchidaceae) by oil collecting bees in Southern Africa. Lindeleyana 4: 164 183. Summerville, Keith S and Thomas O. Crist. 2001. Effects of Experimental Habitat Fragmentation on Patch Use by Butterflies and Skippers (Lepidoptera). Ecology Vol. 82, 5 : 1360 1370. Vinson, Bradleigh S, Gordon W Frankie and J. Barthell. 1993. Threats to the diversity of solitary bees in a neotropical dry forest in Central America. In: Hymenoptera and biodiversity. J. LaSalle and I.D. Gauld, eds. C.A.B. International, Oxon, UK, pp. 53 82. Vinson, Bradleigh S., Gordon W. Frankie and Howard J. Williams. 1996. Chemical Ecology of Bees of the Genus Centris (Hymenoptera Apidae). The Florida Entomologist Vol. 79, 2 :109 129.
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 M39-00123
Evaluando a los rompe vientos como una estrategia de conservacin para las comunidades de polinizadores en los paisajes alterados y fragmentados.
Evaluating windbreaks as a conservation strategy for pollinator communities in altered and fragmented landscapes
In Costa Rica, sustaining human population growth has required fragmentation and
conversion of its Tropical landscape to an agricultural one. Such large scale land use
changes decrease local abundance of species and disrupt important ecological relationships, including species richness and productivity within pollinator communities (Kearns et al. 1997). Pollinators are essential in maintaining stability of plant communities and their dependent consumers. Moreover, in specialized pollinator-host plant mutualisms, the loss of either will reduce or eradicate population of the other. To
counter effects of habitat fragmentation, agricultural communities incorporated
windbreaks into crop fields and pastures as a restorative effort. I evaluate the
performance of windbreaks as repository for pollinators in the San Luis Valley of
Puntarenas, Costa Rica by assessing richness and abundance of both pollinator and insect
communities within three distinct windbreaks and a bordering secondary growth forest.
Pollinator species richness was highest in the forest (H`= 3.72), along with superior
numbers of families and individuals. Two species of non-insect pollinators, hummingbirds Amazilla tzacatl and Hylocharis eliciae, were never seen in the forest but
regularly in two of the three windbreaks. The two sites, one in a shade grown, Coffea
arabica (Rubiaceae) farm and the second, in a pasture, supported comparable species
richness (H` = 2.65 and H` = 2.93), and while a slightly lower amount of individuals was collected in the farm site, it contained nine more families than the pasture. In the same pasture, the third windbreak exhibited lowest diversity (H` = 2.65), and nearly half the number of families and individuals found in the first, neighboring pasture site. Higher pollinator richness and abundance correlated with lower mean temperatures of about 26C and wind speeds in the range of 0.38 m/s 0.9 m/s. Overall, structural attributes, including plant species composition, presence of bodies of water and abiotic conditions, determined a windbreaks capacity to retain pollinators in an altered landscape.
Evalu el funcionamiento de los rompe vientos como deposito para los polinizadores en el Valle de San Luis, Puntarenas, Costa Rica. Determin la evaluacin de la riqueza y abundancia de los polinizadores y de las comunidades de insectos dentro de los tres distintos rompe vientos rodeado del bosque secundario en crecimiento.
Text in English.
Windbreaks, shelterbelts, etc.
Rompe vientos, corta vientos, etc.
Tropical Ecology 2007
Ecologa Tropical 2007
Estrategias de conservacin
t Monteverde Institute : Tropical Ecology