See discussions, stats, and author profiles for this publication at: http://www.researchgate.net/publication/281006149
The Size But not the Symmetry of the Wings of Eulaema nigrita Lepeletier (Apidae: Euglossini) is Affected by Human-Disturbed Landscapes in the Brazilian Cerrado Savanna ARTICLE in NEOTROPICAL ENTOMOLOGY · AUGUST 2015 Impact Factor: 0.85 · DOI: 10.1007/s13744-015-0316-3
4 AUTHORS, INCLUDING: Nelson Silva Pinto
Daniel De Paiva Silva
Universidade Federal da Bahia
Instituto Federal Goiano
13 PUBLICATIONS 19 CITATIONS
13 PUBLICATIONS 48 CITATIONS
SEE PROFILE
SEE PROFILE
Paulo De Marco Júnior Universidade Federal de Goiás 133 PUBLICATIONS 1,293 CITATIONS SEE PROFILE
Available from: Daniel De Paiva Silva Retrieved on: 31 August 2015
Neotrop Entomol DOI 10.1007/s13744-015-0316-3
ECOLOGY, BEHAVIOR AND BIONOMICS
The Size But not the Symmetry of the Wings of Eulaema nigrita Lepeletier (Apidae: Euglossini) is Affected by Human-Disturbed Landscapes in the Brazilian Cerrado Savanna NS PINTO1, DP SILVA2 , JG RODRIGUES3, P DE MARCO Jr3 1
Univ Federal da Bahia, Salvador, BA, Brasil Depto de Ciências Biológicas, Instituto Federal Goiano, Urutaí, GO, Brasil 3 Lab de Ecologia Teórica e Síntese, Depto de Ecologia, Instituto de Ciências Biológicas, Univ Federal de Goiás Campus II, Goiânia, GO, Brasil 2
Keywords Bees, biodiversity crisis, fluctuating asymmetry Correspondence DP Silva, Depto de Ciências Biológicas, Instituto Federal Goiano, Rodovia Geraldo Silva Nascimento, KM 2,5, Zona Rural, Urutaí, GO, Brasil CEP 75790-000;
[email protected] Edited by Fernando B Noll – UNESP Received 24 February 2015 and accepted 10 July 2015 * Sociedade Entomológica do Brasil 2015
Abstract Among other human-related activities, habitat loss and fragmentation are currently ranked as the most important environmental features affecting the persistence of animal and plant populations in landscapes, as well as the maintenance of ecological processes and services. Since these processes are also capable of affecting the ontogenetic development of species inhabiting those landscapes, here we measured the wing veins of male Eulaema nigrita Lepeletier (Apidae: Euglossini) bees in order to evaluate whether the bees sampled in agriculture (AG) areas suffer higher fluctuating asymmetry (FA) than those sampled in Cerrado (CE) areas in the Brazilian state of Goiás. We believe that individuals sampled in CE areas would be less asymmetric than those sampled in AG areas, given a potential higher exposure of these specimens to environmental stresses (mostly pesticides). However, we did not observe a significant trend in the FA measures we obtained, although three wing measures were bigger for bees from CE areas. The lack of significant effects of FA may be related to inherent bionomic features of E. nigrita. For instance, given their high individual dispersal abilities, the individuals we analyzed may have developed in different areas than those where they were sampled. Their generalist feeding behavior may also have given them a higher resistance to environmental perturbations, buffering the normal development of immatures even in areas with local high environmental stress. Nonetheless, higher death rates of individuals from anthropic areas may also have killed the developing immatures of E. nigrita before they reached adulthood consequently equalizing our sampled estimates.
Introduction Human-related activities are responsible for great environmental changes nowadays with inherent pervasive effects on biodiversity (Sala 2000, Chapin et al 2000). Processes such as habitat loss and fragmentation, the introduction of exotic species, and the deposition of nitrogenous compounds are ranked as the main causes of the current and future biodiversity crisis (MEA 2005, Tylianakis et al 2008). Eventually,
such impacts affect important environmental features, which are important determinants for the biodiversity survival (e.g., decreases in reproductive output, genetic quality, demographic rates, loss of dispersal abilities, changes in species richness and abundance, as well as trophic cascading effects, among others; Fahrig 2003), as well as important services like pollination. Bees are important pollinators to the great majority of flowering plants (Klein et al 2007, Winfree et al 2011), and
Pinto et al
euglossine bees are one of the most important pollinators in Central and South America (Tonhasca et al 2003). Eulaema nigrita Lepeletier (Apidae: Euglossini; sensu Silveira et al 2002) bees are assumed to be tolerant to environmental disturbance (Peruquetti et al 1999, Tonhasca et al 2003), showing lower abundances in big-sized fragments or conserved forested areas (Brosi 2009, Tonhasca et al 2003). Furthermore, males and females present smaller body sizes in fragments surrounded by crop activities (Peruquetti 2003), that may affect the developmental stages of exposed immature individuals up-to their adulthood. In general, bioindicator studies only consider demographic and community analyses of the local available species pool exposed to environmental stressors to assess their effect on the biotic component (McGeoch 1998). Nonetheless, lowcost alternatives, that only consider the individual level, are also available and are capable of providing quantitative results of the direct effects of environmental disturbances. Indexes of ontogenetic developmental instability during immature stages of the analyzed species, such as fluctuant asymmetry (FA), are cheap methods to evaluate the effects of environmental stressors related or not to human activities (Hogg et al 2001, Sanseverino & Nessimian 2008, De Block et al 2008, Silva et al 2009) and are not related to evolutionary processes. In these analyses, the amount of morphological changes of mature individuals affected by a stressing event during their immature stages may be used to evaluate individual developmental instability in a given area (Weller & Ganzhorn 2004), with direct relationship to their adaptability to local environmental features. For instance, wing shape and size directly influence the way flying animals disperse in their surrounding environment, and symmetric individuals are expected to fly more effectively due to better aerodynamics when compared to the asymmetric ones (De Block et al 2008). Therefore, the existence/occurrence of high levels of FA in wing shape and size may directly affect foraging activities, escape from predators, and mating behaviors (Córdoba-Aguilar 1995, Bots et al 2009, Stewart & Vodopich 2013). Additionally to causing FA, environmental stressors during the ontogenetic development of the focus species may cause smaller wing sizes in individuals that developed near disturbed areas, when compared to those from preserved areas. In order to cope with the environmental stressor and continue their development, some physiological triggers may act during the species ontogenetic development, resulting in smaller adult individuals. For instance, in laboratory conditions, stressors (e.g., high temperatures) have proved to cause an increased development of the studied individuals and a smaller final body size when compared to individuals maintained under low stress conditions (Kanegae & Lomônaco 2003, Chang et al 2007). Pinto et al (2012) found that the specimens of Argia tinctipennis Selys living in
anthropized areas had smaller wing sizes than those living in preserved sites. Finally, specifically for hymenopterans, food quality and quantity are especially important to determine the final body size of both males and females (Roulston & Cane 2000). Therefore, the amount of human-disturbed areas determining the quality of the landscapes surrounding the bees nest may eventually determine the body size of the adult bees. Once the ontogenetic development of the biodiversity inhabiting human-affected landscapes may be directly affected by habitat loss and fragmentation, here we investigated whether the ontogenetic development of E. nigrita bees sampled near intense-used agricultural (AG) vs. Cerrado (CE) areas was affected by these processes. As our general expectation, we believed that individuals sampled in CE areas would not show fluctuating asymmetry in their wings, whereas those sampled in AG areas would show higher amounts of asymmetry. Additionally, we also compared the mean size of the measures of the wings of individuals from both CE and AG areas. Since the exposure to stressing events during their development was expected to be higher in AG areas, we expected that individuals from these areas would show smaller wing size measures than those sampled in CE ones.
Material and Methods The focal Euglossini species Generally, Eulaema bees are large-bodied species with 20– 30 mm of body length, dense presence of bristles, and integument not showing metallic coloration hue in their heads and mesosoma (Silveira et al 2002, Oliveira 2006). These species build parasocial nests in tree cavities, where several females lay their eggs within the same batch of brood cells, which are generally build with mud and feces mixed up with glandular substances and/or resins (Zucchi et al 1969, Santos & Garófalo 1994, Silveira et al 2002). It is also one of the most common species of Eulaema in degraded areas with open areas and/or intense influence of anthropic activities, being considered as a bioindicator of degraded areas (Peruquetti et al 1999, Neves & Viana 1999, Tonhasca et al 2002) and/or open habitat, like those within the Cerrado savanna (Silva & De Marco 2014 and references therein). This species is abundant in open areas such as the Cerrado (Silva & De Marco 2014 and references therein), as well as in degraded/anthropized landscapes (Tonhasca et al 2003). Finally, E. nigrita shows excellent flight and dispersal ability (Ackerman 1983, Raw 1989), that may explain its protection against environmental degradation (Silva & De Marco 2014).
Lack of Asymmetry Differences on Eulaema nigrita Bees from Cerrado
Study areas and sampling methods Individuals of E. nigrita from this study were sampled in areas of the Brazilian Cerrado (CE) and in intensely used agricultural (AG) areas from the state of Goiás, in the core area of this Brazilian biome during the late wet seasons of 2011 (February to April) and 2012 (March to April). Since males of E. nigrita may have different body sizes from one season to the other (Peruquetti 2003), the standardization of the same sampling season within different years decreases the effect of varying body sizes in our FA analyses. This biome has high species richness and endemism and currently faces a fast process of native areas conversion to soybean and corn croplands, as well as pastures for extensive cattle raising (Klink & Machado 2005, Carvalho et al 2009). The most recent estimates point out that 39–55% of the original Cerrado cover was converted to areas deeply involved with human-related activities (Klink & Machado 2005, Sano et al 2008). The overall sampling methods employed here may be found in Silva & De Marco (2014). Nonetheless, the selection of the sampling areas occurred considering macro- and micro-regional scales. At first, we used the land use classification of Cerrado produced by Sano et al (2008) and divided the state of Goiás into 25×25 km quadrats. Then, we used FRAGSTATS v3.3 (McGarigal & Marks 1995) to calculate the amount of natural vegetation and croplands available within each quadrat. We excluded boundary quadrats and those having more than 95% or less than 5% of both agriculture and natural areas. We generated a gradient between the remaining proportion of native Cerrado vegetation and the isolation among these areas within the remaining quadrats. Later, we selected 11 quadrats that represented the overall fragmentation pattern observed for the state. In the microregional scale within these quadrats, we sampled 33 areas surrounded by agriculture (AG areas) and 17 areas surrounded by native Cerrado (CE areas). A scheme for this whole process may be found in Silva & De Marco (2014). All sampling took place near streamlets, since euglossine bees are intimately associated with humid forested areas (Dressler 1982, Roubik & Hanson 2004, Moura & Schlindwein 2009), mainly in arid areas in the Neotropical region, using riparian areas as dispersal corridors (Moura & Schlindwein 2009). In each sampled area, we delimited a 250-m transect, and at every 50 m, we set up a sampling station, totaling six stations in each area. Each station was composed of five conventional scent traps made of PET bottles with three entrances to sample euglossine bees (Campos et al 1989), which were installed at a height of 1.5 m and separated from one another by 5 m. A wooden rod with cotton swabs embedded with an aromatic scent (vanillin, eucalyptol (cineol), eugenol, methyl-salicylate, or methyl-cinnamate) was provided within each scent trap. For each sampling station, there was one scent trap for each specific scent.
The scent traps remained in the field for approximately 48 h. In total, we sampled 238 E. nigrita bees (167 from CE and 71 from AG areas). Sampled specimens were preserved in ethanol and later pinned, identified, labeled, and deposited in the Entomological Collection of the Theory, Metacommunity, and Landscape Ecology lab from Universidade Federal de Goiás. Wing measures of E. nigrita At the lab, we removed the fore and hind wings of both left and right sides of the sampled specimens using dissection entomological scissors. Since the wings of E. nigrita are dark, what makes the wing venule visualization difficult, we diaphanized their wings with hydrogen peroxide in open air for 24 h. After such procedure, we affixed the wings with visible venules to a white paper sheet with transparent adhesive tapes (Scotch Brite® 3M). Later, we digitized such sheets using a Hewlett-Packard® desktop scanner. The digitized wings individually identified to every single specimen were measured. We measured 11 wing venules on both fore and hind wings of every sampled individual with the software tpsDIG 2.1 (Rohlf 2006). Of these 11 venules, eight were from the fore wings (measurements 1–8; Fig 1), while three were from the hind wings (measurements 9–11; Fig 1). We performed two different measurements in different days to evaluate the measurement error rate, as proposed by Palmer & Strobeck (1986). Fluctuating asymmetry, measurement errors, and statistical analyses We followed the protocol proposed by Palmer & Strobeck (1986) to measure the wings of E. nigrita. We chose to subtract the measurements of the right side from the very same measurements made in the left side of each wing (fluctuating asymmetry=FA). Such index directly evaluates the deviation of a bilateral perfect symmetry, where symmetry is always equal to zero. We used a factorial mixed model ANOVA, using “individuals” as a random factor and “side” as a fixed factor. In this model, it is possible to verify whether the measurement errors are significantly higher than the observed FA measurements through an analysis of the mean squares for each factor (sides and individuals) and for the measurement error (Palmer & Strobeck 1986). We used a t test of a sample against zero to exclude the possibility of other asymmetry (directional symmetry and antisymmetry) and a Shapiro-Wilk test to evaluate whether the data distribution was bimodal (Palmer & Strobeck 1986). In the case we detected values different than zero in the t test results, we built scatterplots of right vs. left measures (in the ordinates and abscissas axes, respectively). Such procedure was necessary because we observed few categories of
Pinto et al
Fig 1 Representation of the measured veins in both fore (measures 1–8) and hind (measures 9–11) wings of Eulaema nigrita bees sampled in this study.
FA, which may represent false positives (leptokurtic distribution; Palmer 1994). In this visual analysis, an aggregation of cases is an indicative directional asymmetry. On the other hand, two aggregations of cases are an indicative of antisymmetry. Therefore, the expected pattern for the occurrence of FA in the measured individuals would be a random distribution of cases in the scatterplot. Here, we assumed the assumptions from Babbitt (2006) that considers that the distribution of FA cases does not need necessarily to be adjusted to Gaussian distribution, as observed in other studies (e.g., Souto et al 2011). We used t tests to evaluate whether the FA levels obtained for the specimens sampled in the AG areas were higher than those observed in CE areas (Zar 2010), where the FA wing measurements of each side were the dependent variable used in the tests. Additionally, we also compared the mean size of the measurements of each individual, from each area category (AG vs. CE) using independent t tests. Since the measurement errors between the first and the second measures were too low (Table 1), we used the mean of the mean of each measurement from each individual from each different area category as our dependent variable to evaluate the FA differences between bees. If the assumption of homogeneity of variances was not met, we used t test with separated variances (Zar 2010). A summary of the analytic methods employed in this study may be found in Fig 2.
Results Except for measurements M1 and M3, the measurement error of the wings was smaller than the effects of both individuals and side terms (Table 1). In the further analyses, we excluded both M1 and M3 as
possible measures suffering FA in the sampled specimens of E. nigrita. Since, the differences associated to the side effect are not explained by chance (Table 1), the observed wing asymmetry of E. nigrita is not considered as directional. However, none of the variables showed normal distribution (Table 2). We evaluated both the skewness and kurtosis of the distribution of our measurement data. No variable showed asymmetry estimates higher or lower than +1 or −1, respectively, but they showed values indicating leptokurtic curves, what suggests that the distribution curve is symmetrical with many values near the mean value.
Table 1 Error estimate of the measurements in both fore and hind wings of both left and right sides of the Eulaema nigrita specimens sampled in both agriculture (AG) and Cerrado (CE) areas from the state of Goiás [df=236 (treatments: 1; residuals 235); * p
