Research Paper in Ecology Lab

ECOLOGY, BEHAVIOR AND BIONOMICS

Eucalyptus Edge Effect on Quercus-Herbivore Interactions in a Neotropical Temperate Forest

C HERNÁNDEZ-SANTIN1, M CUAUTLE1 , M DE LAS N BARRANCO-LEÓN2, J GARCÍA-GUZMÁN1, El BADANO2, F LUNA-CASTELLANOS1

1Depto de Ciencias Químico Biológicas, Univ de las Américas Puebla, Cholula, Puebla, Mexico 2División de Ciencias Ambientales, Instituto Potosino de Investigación Científica y Tecnológica, San Luis Potosí, Mexico

AbstractKeywords

Quercus , herbivory, edge effect, Lepidoptera caterpillars

Correspondence M Cuautle, Depto de Ciencias Químico Biológicas, Univ de las Américas Puebla, Cholula, Puebla, Mexico; mcuautle2004@ hotmail.com

Edited by Martin F Pareja – UNICAMP

Received 18 June 2018 and accepted 26 April 2019

* Sociedade Entomológica do Brasil 2019

Fragmentation leads to the formation of edges between habitats, which in turn changes biotic and abiotic factors that might influence herbivory or plant-herbivory interactions. The aims of this study were to describe the herbivory community associated with oak (Quercus) and to determine the effects of proximity to a Eucalyptus edge and season on insect herbivory. We selected three forest sites that were subsequently divided into three quadrants located at different distances from the Eucalyptus edge: edge (0 m), intermediate (30 m), and oak forest interior (60 m). We randomly selected 10 oak trees per quadrant and conducted monthly surveys, during the dry and rainy season (from February to October 2010), where we quantified leaf area and the percentage of herbivory. These were analyzed using linear mixed models, with distance and season as fixed factors and individual and site as random factors. The primary oak herbivores were Lepidoptera caterpillars. We found that herbivory increased away from the edge but just during the rainy season, although higher herbivory levels were found during the dry season. These results seem to be related to a specialist community of herbivorous associated to the Quercus. This study emphasizes the importance of considering border effect, especially within Natural Protected Areas to establish strategies to improve and maintain native oak forest and the biodiversity of its Lepidoptera herbivorous community.

Introduction

Landscape modification due to anthropogenic activities (e.g., land conversion to agricultural or livestock) has resulted in habitat fragmentation, one of the major threats for forest conservation (Buckley 2000, Franklin et al 2002). Fragmentation is defined as the disruption or breakdown of large vegetation patches into smaller ones resulting in a dis- continuity of resource distribution that affects species occu- pancy, reproduction, and/or survival (Franklin et al 2002). One of the important features of this phenomenon is an increase in edge length relative to the forest area, particular- ly in small habitat fragments (Laurance 1991, Laurance &

Yensen 1991, Murcia 1995, Laurance et al 2007, De Carvalho Guimarães et al 2014). Edges are defined as bound- aries between distinct patches types, so edge identification depends on how patches are defined (Ries et al 2004, Wirth et al 2008). Both physical (light, humidity, solar incidence; Buckley 2000, Murcia 1995, Williams-Linera et al 1998) and biotic conditions (e.g., plant-insect interactions, species and resource distribution) are strongly altered in forest edges relative to interior (Lidicker 1999, Bogaert et al 2001, López 2004, De Carvalho Guimarães et al 2014), and these changes are referred to as edge effects (Wirth et al 2008). The chang- es in the physical environment are followed by changes in forests structure and species composition and abundance

Neotrop Entomol https://doi.org/10.1007/s13744-019-00694-5

along the forest edges (Ries et al 2004, De Carvalho Guimarães et al 2014). Finally, changes in species composi- tion may result in modified species interactions (De Carvalho Guimarães et al 2014).

The ecological consequences of edge effects have been mostly studied in terms of effects on biodiversity and species abundance and distribution (Lidicker 1999, Bogaert et al 2001, López 2004, Medinaceli et al 2004, Connor 2006), whereas studies on edge effects on species interactions such as herbivory have received relatively less attention (but see, Baraza et al 2004, Bossart and Opuni-Frimpong 2009, Liang & Stehlik 2009, Rossetti et al 2014, De Carvalho Guimarães et al 2014). Herbivory, defined as the consumption of plant tissue by animals (Strauss & Zangerl 2002), is the most wide- spread plant-insect interaction and represents an important determinant of vegetation structure (Medinaceli et al 2004). There are two types of herbivory: external herbivory occurs when herbivores feed on leaves, flowers, and seeds, whereas internal herbivory takes place under the plant’s surface and is caused mainly by insects that cause galls or mines (Strauss & Zangerl 2002). A meta-analysis (31 studies) of the effects of fragmentation on herbivory, performed by De Carvalho Guimarães et al 2014, found a significant increase in insect richness (65%), insect abundance (14%), and plant herbivo- rous rates (74%) on edges, compared with vegetation inte- riors. They also found that herbivorous rates were higher when herbivorous were chewers of the Orthoptera and Lepidoptera orders. AccordingtoDe CarvalhoGuimarães et al 2014, edges can be considered postdisturbance recovering sites, in which plant species number and/or densities incre- ment, representing extra resources for generalist herbivo- rous. This study is in accordance with the revision (55 studies) made by Wirth et al (2008), who found that 82% of the studies reported that edge promotes a positive effect on herbivorous response (abundance, species richness).

In the study area, the native vegetation is oaks forest and the main herbivores are Lepidoptera caterpillars. Oaks (Quercus spp.), from the Fagaceae family, are a group of trees and shrubs distributed throughout the Northern hemi- sphere including Europe, Asia, and North America (Vázquez et al 2004). In Mexico, oak-dominated forests cover 5.5% of the country’s land total surface and include 161 Oak species of which 109 are endemic (Arizaga et al 2009). Based on this, Mexico is considered a center of diversification of oaks in the American continent (Vázquez et al 2004). In the Neotropics, oaks are widely distributed, ranging from the north of the Mexican Transition Zone down to the south of the Colombian Andes, and it includes the greatest oak species diversity re- gion in the mountains of southern Mexico (Rodríguez-Correa et al 2015). The main insect herbivores on oak species belong to Lepidoptera and Hymenoptera which feed on leaf tissue during the larval stage (Summerville et al 2003), and Lepidoptera in particular represents a dominant group of

herbivores on oak by causing up to 88% of the external herbivory on these tree species (Hochwender et al 2003, Williams-Linera & Herrera 2003). Previous work with oak herbivory has found that it is influenced by abiotic factors such as light availability (Baraza et al 2004, Liang & Stehlik 2009), as well as biotic factors such as the age and size of the plant (Medinaceli et al 2004), plant tissue type (Baraza et al 2004), and tissue quality (Murakami et al 2005).

The state park “Flor del Bosque” is a Natural Protected Area (SEMARNAT-CONANP 2013–2018), located in Puebla (Mexico), which has an oak forest. However, in recent de- cades, increasing demand for firewood and clear cutting to establish Eucalyptus spp. plantations have resulted in dra- matic effects of fragmentation in the oak forests (Costes- Quijano et al 2006). The given condition in the study area is similar to other Natural Protected Areas that present some degree of the environmental deterioration caused by land use change, pollution, and habitat fragmentation (Cruz- Elizalde et al 2015, Ochoa-Ochoa et al 2009). Given the pres- ent condition of the study area, understanding Eucalyptus edge effects on species interactions occurring within oak for- est stands is of vital importance for conserving insect diver- sity and oak native populations. The questions addressed in this study included the following: (1) What are the primary external herbivores of the Quercus species in a neotropical temperate forest? (2) What is the Quercus-related effect of proximity to Eucalyptus edges on external herbivory? (3) Does external herbivory varies among seasons (dry and rainy)? We hypothesized that insect herbivory associated with Quercus would decrease away from Eucalyptus spp. edges. It is expected that external herbivory will be higher at the Eucalyptus spp. edge and will decrease gradually to- ward the forest interior. In the study zone, most of the but- terflies reproduce in the rainy season (Barranco-León et al 2016), so we expect higher herbivory in this season.

Materials and Methods

Study area

The Natural Protected Area “Flor del Bosque” is located in the state of Puebla, Mexico (between 19°00′00″ and 19°01′ 50″N, and 98°20′35″ and 98°20′53″W). The study area has sub-humid temperate weather with rains during the summer (June–September) while a markedly dry season occurs during the rest of the year (Cwb in Köeppen classification), an an- nual mean temperature of 18°C, and mean annual precipita- tion of 800 mm. Relative humidity ranges from 42 to 67% depending on levels of precipitation. The Natural Protected Area covers a total of 664 ha of which 255 ha (41%, of the total area of the protected area) is constituted by oak forest, 9.3 ha (1.5%) is covered Eucalyptus plantation, and the

Santin et al.

remaining 400 ha is covered mostly by pasture (Costes- Quijano et al 2006).

Oak forests include 10 oak species (Badano et al 2010). In general, forest patches found within the park are highly dis- turbed and trees tend to be small in stature and highly branched. The dense vegetation present in these areas gen- erates more humid conditions and lower light availability, and the dense forest cover prevents rainfall from reaching the ground directly (Barranco-León et al 2016).

Eucalyptus patches are represented by two species, Eucalyptus camaldulensis Dehnh and Eucalyptus globulus Labillardière, which were introduced more than 45 years ago. In contrast to oak forest patches, Eucalyptus stands allow much higher levels of rainfall to reach the soil (81%) and are characterized by greater light availability (Barranco- León et al 2016).

Herbivore community

The main external herbivores that were observed feeding on Quercus species were Lepidoptera. Internal herbivory was also observed, done by insects that cause galls or mines. In this study, we refer to just external herbivory caused by Lepidoptera. From September 2009 to October 2010, Lepidoptera caterpillars that were seen feeding on Quercus spp. were collected and taken to the laboratory for observa- tion until metamorphosis. In each case, the host tree species was recorded. Caterpillars were reared at the Entomology Laboratory of the Universidad de las Américas Puebla and maintained at room temperature (20°C) and 60% relative humidity. Within the rearing room, caterpillars were individ- ually placed inside a 2.5-L plastic container or a plastic bag. Old leaves fed to caterpillars were periodically replaced by new leaves. Date of pupation was recorded for each cater- pillar. When necessary, the containers were filled with a mix- ture of dry leaves and dirt for the caterpillars to dig in and pupate. After metamorphosis, adults were sacrificed for identification. All Lepidoptera were identified to the mini- mum taxonomic level possible using the proposed identifica- tion keys from Borror and White (1970), Borror et al (1992), Marshall (2006), Powell & Opler (2009), and mounted spec- imens were revised by Ph. D Hantke and determined to species. Oaks from which caterpillars were collected were identified whenever possible to species using the Rzedowski and de Rzedowski (2001) key for oaks.

Study design

Due to the fragmentation process that has occurred in the park, the oak forest patches are of different size and border Eucalyptus plantation or pastures. We selected three oak forest patches (i.e., hereafter “sites”) that were bordering a Eucalyptus plantation, to standardize for Quercus-Eucalyptus

border. In each site, we delimited three 20 × 20 m quadrants. The quadrants were established at the edge (0 m from the Eucalyptus stand), 30 m from the edge (intermediate zone), and 60 m from the edge (forest interior).

Within each quadrant, we randomly selected five trees each of Quercus castanea Née and Quercus obtusata Bonpland. These two oak species were chosen because they were abundant and/or its determination to species was pos- sible. Determination of Quercus spp. was just possible for the site two and three. The diameter at breast height of the oak individuals selected range from 28 to 990 cm, with an aver- age of 365 cm. For each tree, we harvested three branches (maximum branch height of 5.5 m) and transported branches to the laboratory for quantification of herbivory. This sam- pling procedure was conducted during the first week of each month, from February to October of 2010, which included months of the rainy (June, July, August, and September) and dry season (February, March, April, May, and October).

Leaf area and herbivory measurement

For each sampled branched per tree, we numbered 25 leaves and randomly selected 15 to record the percentage of exter- nal leaf herbivory. A total of 12,150 leaves were sampled, 1350 per month. To estimate herbivory, we calculated leaf area by scanning each leaf and subsequently digitally recon- structed the leaf (when there was tissue lost to herbivores) using Paint software n.d. (version 6.1, Microsoft ®). This re- sulted in estimates of area per leaf, one before and the other after herbivore damage. We measured leaf area using the program “Medición de objetos” version 4.2 (object ’s mea- suring) (Ordiales-Plaza 1999–2000). The percentage of leaf area consumed was estimated as actual or observed leaf area/reconstructed leaf area prior to defoliation. For statisti- cal analysis, we averaged leaf area and herbivory values across leaves of all three branches sampled per tree, sepa- rately for each survey.

Statistical analysis

We performed a linear mixed model analysis model testing for the effects on herbivory (proportions) of distance to the edge (border, 30 m, and 60 m to the edge) and season (rainy and dry) as fixed effects and site and tree individual as ran- dom factors, this last random factor accounts for the repeat- ed measures in time. In the mixed model, one or more ran- dom effects are added to the fixed effects. These random effects essentially give structure to the error term “ε”. In our study design, this resolves the non-independence that stems from having multiple responses within the same site (Winter 2014). Response variables involving proportions are ade- quately evaluated using a binomial distribution (Crawley 1993). All analyses were performed with the statistical

Eucalyptus Edge Effects on Quercus-Herbivore Interactions

software of the package “glmer” of the software R ver. 3.2.2 using binomial distribution (R Core Team 2015). Assumptions of normality and homoscedasticity were evaluated with Q-Q plot and fitted vs residuals. Significance of the factors was evaluated using a likelihood ratio test which compares the likelihood of two models: the model with the factor of inter- est and the model without the factor of interest. Whenever a factor was significant, we performed post hoc mean con- trasts (Tukey method). A similar analysis was performed but considering just the two sites in which the Quercus spp. were determined. The response variable was herbivory pro- portions, distance to the edge (border, 30 m, and 60 m to the edge), season (rainy and dry), and Quercus species (Q. castanea and Q. obtusata) as fixed effects and site and tree individual as random factors.

Results

Herbivore community

A total of 70 caterpillars distributed across 24 morphospecies were sampled from the oak trees sampled. Of these 70 spec- imens, 43 pupated, and of these 28 completed their meta- morphosis and developed into adults. The 24 morphospecies sampled comprised 10 families of caterpillars, namely: Arctiidae, Geometridae, Hesperiidae, Limacodidae, Lymantriidae, Lasiocampidae, Megalopygidae, Noctuidae, Riodinidae, Saturniidae. In addition, we also sampled one morphospecies of beetle (Coleoptera: Coccinellidae). The species that were identified were Emesis zeal Butler (Riodinidae); Anisota assimilis Druce, Automeris cecrops Boisduval, Automeris randa Druce, Paradirphia lasiocampina R. Felder and Rogenhofer, (Saturniidae); Artace cribraria Ljungh, Gloveria olivacea H. Edwards (Lasiocampidae); Trosia obsolescens Dyar (Megalopygidae); Prolimacodes badia Hübner (Limacodidae). These species were identified or revised by Ph. D. Hantke. Of the collected morphospecies, 20 were found feeding on Q. castanea, four on Q. obtusata, three on Q. laeta Liebman, and one on Q. laurina Bonpland. Only three morphospecies were reported feeding on more than one oak species.

Eucalyptus edge and season effect on leaf herbivory

We found significant effects of distance to the edge (X2 = 7.97, gl = 2, P = 0.018) and season (X2 = 6320.2, gl = 1, P < 0.001) on the percentage of leaf herbivory, significant two-way interactions among distance and season were found (X2 = 530.7, gl = 2, P < 0.001). Specifically, we found that the amount of herbivory at the forest edge (0 m) was significantly lower than that found at the intermediate dis- tance and the interior (t > 2.1, P < 0.02). With respect to

season differences, we found that in the rainy season the amount of herbivory was significantly lower than that in the dry season. A significant interaction between distance and season, on herbivory, was found (X2 = 3.353, gl = 2, P = 0.187, Fig 1A). Interaction effect showed that the amount of herbivory at the forest edge (0 m) was significantly lower than that found at the intermediate distance and the interi- or, but just during the rainy season (Tukey method, P < 0.001). During the dry season, the same tendency was found, but this was not significant (Tukey method, P > 0.30, Fig 1A).

Analyses based upon the comparison of Q. castanea and Q. obtusata for sites 2 and 3 showed that the percentage of herbivory for Q. obtusata (6% ± 0.01) was significantly higher than that for Q. castanea (4% ± 0.02) (X2 = 9.59, gl = 1, P < 0.002). The effects of the season (X2 = 3298.2, gl = 1, P < 0.001) and the interaction among these two factors were significant (X2 = 28.52, gl = 2, P < 0.001). For the interaction, not significant differences were found among the herbivory percentage of Q. castanea during the season with higher herbivory (i.e., dry) and Q. obtusata during the season with lower herbivory (Tukey method, P = 0.452). The other com- parisons were significant (Tukey method, P < 0.007, Fig 1B).

Discussion

This study describes the community of herbivores associated with the oak species of native forest in the Neotropics and evaluates the effect of the distance to the edge and season on herbivory. The oak herbivorous community was mainly composed of lepidopterans, and we found greater herbivory at the interior of the forest, which may be related to specific microclimatic conditions at the interior of the forest or to the presence of a specialist Lepidoptera community. However, this pattern was significant just during the rainy season, al- though the higher amount of herbivory was found during the dry season. Contrary to what was expected, herbivore levels were not higher at the Eucalyptus edge, and this result may be explained by the idiosyncratic characteristics of the lepi- dopteran community associated to the oak forest (i.e., spe- cialist species). We discuss in detail the results of this study.

All of the families to which the insect specimens were sampled, except Coccinellidae, are constituted exclusively by Lepidoptera herbivorous species. In addition, the Noctuidae family has been previously reported as one of the most important groups associated with Quercus species (Monreal et al 1992, Hochwender et al 2003, Extremera et al 2004). Of lesser importance but also previously reported for oaks are Lymantriidae (Monreal et al 1992, Extremera et al 2004) and Lasiocampidae family (Monreal et al 1992). The 24 morphospecies recorded, which belonged to 10 families, in- dicate a high diversity of lepidopteran herbivores that feed

Santin et al.

and possible depend in an important way of the Oak species in the zone. This diversity agrees with the high diversity found by Barranco-León et al (2016) for diurnal lepidopteran species, in the same study zone, who found 91 species contained in the six taxonomic families of diurnal lepidop- terans. Barranco-León et al (2016) point out that the diversity found in the park reserve is greater than the one reported in larger forest ecosystems in Mexico.

We found that levels of herbivory on oak plants decreased significantly at the edge of Eucalyptus stands relative to plants located in the quadrants inside forest patches. This pattern indicates that Lepidoptera herbivores are negatively affected by proximity to a border Quercus-Eucalyptus. This result contrasts with the findings of De Carvalho Guimarães et al (2014), who found higher herbivorous levels at the edge than at the interior of temperate forests. They attributed this result to an increase in plant abundance in edge habitats, as the key mechanism, leading to an increase in herbivorous levels. However, De Carvalho Guimarães et al (2014) point out that this explanation would apply to a generalist herbiv- orous community; meanwhile, a specialist herbivorous com- munity would not be favored by an increase in this plant abundance as much of them would belong to unpalatable plant and/or replace their local host. In this study, the

Lepidoptera herbivores probably are oak specialist, for which the Eucalyptus edge represents a degraded habitat of lower resource quality. Santos & Benítez-Malvido (2012) found a similar result to ours; they evaluated the caterpillar herbivory damage on Heliconia latispatha, and they found that herbiv- ory levels diminished from 4.2% in forest gaps to 0.5% on road edges. They attributed this result to changes in micro- climatic conditions, host quality (higher physical vs. chemical defense on edges), and local abundance of herbivores (lower caterpillars abundance, due to higher predation pressure and/or dispersal limitation on edges).

Microclimatic conditions could explain the differences in herbivory levels between the edge and the forest interior. In the study zone, Barranco-León et al (2016) reported higher temperatures and photosynthetic active radiation (light) in the Eucalyptus plantation than in the oak forest; meanwhile, the humidity was higher in the oak forest. We argue that Quercus herbivores are not favored by greater light availabil- ity; this result is similar to that reported by Baraza et al (2004) with Quercus seedlings, who found that plants grow- ing in the shade had lower levels of chemical defenses than did sun-grown plants, making them more attractive to her- bivores. Related to this pattern is the study of Connor (2006) who found that survival rates for species of Lepidoptera de- creased with increasing light availability probably due to dif- ferences in foliage quality (Connor 2006). In our study, mild temperatures and higher humidity, at the interior of the oak forest, especially during the Lepidoptera reproduction period (rainy season), could favor them preventing egg desiccation and promoting larvae development (Barraco-Leon et al 2016). The negative effect of Eucalyptus edge on Lepidoptera herbivory community could be accentuated more in the reproductive period of the Lepidoptera and ex- plains why is in this season that we found a significant effect of the distance to the edge. However, our findings disagree with other studies reporting that herbivory is favored by increasing light availability at forest edges (Lincoln & Mooney 1984, Moore et al 1988, Basset 1991, but see Medinaceli et al 2004, Connor 2006, Rossetti et al 2014, Pereyra et al 2015).

The degree of host specificity of the caterpillar species associated with Quercus could also explain the herbivorous pattern observed. The comparison of herbivory on Q. castanea and Q. obtusata indicated that the latter was more heavily attacked. We speculate that Q. castanea is of higher quality (fewer defenses or higher nutrient content) than Q. obtusata and is thus preferred by the sampled spe- cies of Lepidoptera. Quercus obtusata might have specialist Lepidoptera that caused more damage than the greater but less specialist Lepidoptera community associated to Q. castanea. Similarly, Maldonado-López et al (2016) found that gall wasp richness depended on the Oak species considered.

0

2

4

6

8

10 A

B

0 m 30 m 60 m

H er

bi vo

ry p

er ce

nt H

er bi

vo ry

p er

ce nt

Distance from Eucalyptus edge

Dry Rainy

0 1 2 3 4 5 6 7 8 9

Q. castanea Q. obtusata Quercus species

a

c

a

b

a

b

a

b

c

a

Fig 1 A Edge effect on the percent of leaf area lost to Lepidoptera herbivores on oak for the dry and rainy season (mean herbivory percent ± se) (Quercus spp.). Herbivory percent was significantly lower on edge in comparison with the oak forest interior (30 m and 60 m), but just during the rainy season. B Mean herbivory percent for the two Quercus species considered (Quercus castanea and Quercus obtusata, sites 2 and 3). Different letters (a, b, c) indicate significant differences (Tukey method, P < 0.001).

Eucalyptus Edge Effects on Quercus-Herbivore Interactions

The idea of specificity of the caterpillar species associated with Quercus is in accordance with the findings done by Barranco-León et al (2016), who studied lepidopteran com- position in the area and considered the three different types of vegetation within the reserve (oak forest, grassland, and Eucalyptus plantation). They found that lepidopteran diurnal species, mainly specialist species, were associated with oak forest; meanwhile, no positive correlation (i.e., the observed frequency of the registered lepidopteran species were not bigger than that of expected by random) were found be- tween the Eucalyptus plantations and any lepidopteran spe- cies, even though a number of butterflies species were re- corded in this type of vegetation. They concluded that al- though Eucalyptus plantations could be used as corridors by the butterflies, this type of vegetation was not essential for preserving native lepidopteran species.

We also found evidence of seasonal variation in the levels of herbivory, as indicated by significant lower herbivory dur- ing the rainy season. In the zone, Quercus trees shed their leaves in March, so the lower levels of herbivore found dur- ing the rainy season correspond to herbivory on the new leaves; meanwhile, a greater herbivore level damage was found during the dry season, which corresponds to herbivory on older leaves. This contrasts with other communities in which young leaves feeders represent an important compo- nent of the Lepidoptera community (Forkner et al 2008). However, these results are consistent with the lepidopteran seasonality found for the lepidopteran community in the area, for which it is reported a higher diversity of forest oak Lepidoptera specialists as well as generalist, during the dry season (specialist 33 species in the rainy season vs 39 species during the dry season, generalist 20 species vs 30, respec- tively (Barranco-León et al 2016)). Similarly, Maldonado- López et al (2016) found a richer gall wasp community during the autumn than during the spring in an oak forest in Mexico, associated with a higher quantity of leaves in this season.

This study highlights the negative effects of the proximity of an oak forest edge to Eucalyptus spp. stands. Border effect could be resulting in higher levels of herbivory inside forest patches, and this in turn could lead to reductions in Quercus spp. growth and survival. The effect of insect herbivory on plant growth and survival will depend on the level of herbiv- ory, the leaf value (e.g., old vs new leaves), and the host condition (Schowalter et al 1986, Rossetti et al 2014). Previous work has shown that levels of herbivory over 20% cause a decrease in reproductive success for Q. alba Linneo and Q. robur Linneo (Crawley 1985, Hochwender et al 2003). Given that forest edge effects will likely become more impor- tant with increasing levels of fragmentation, we argue that further work is necessary in order to uncover the mecha- nisms by which fragmentation and the edge effects shape Quercus-herbivore interactions and the consequences of such changes for oak survival and recruitment, as well as

for insect abundance and diversity. Future work should ad- dress the consequences of changes in herbivory due to edge effects on oak survival rate and seed germination.

We found evidence for edge effects on insect leaf damage associated with oaks. We found that leaf herbivory was low- er at the forest edge relative to the interior, a finding which is consistent with the scenario of a rich site neighboring a de- graded one (Ries & Sisk 2004). However, this effect was just found in the rainy season, which may be related to the re- production period of the principal herbivorous found: larvae of Lepidoptera, which probably prefer the mild temperatures and higher humidity, at the interior of the oak forest. Some of the caterpillar morphospecies sampled could be special- ized on Quercus spp., although further research is necessary to characterize the degree of dietary specialization of these herbivores. Further work is also needed to determine if ob- served levels of herbivory are affecting oak survival, recruit- ment, and reproduction. Research should also focus on de- signing strategies aimed at reducing the negative effects of forest edges on oak populations. Some of these strategies could be aimed at re-establishing and maintaining balanced plant-herbivore interactions as well as re-introducing native oak species. The prevalent conditions in this park are similar to those faced by most protected areas: a natural conserved is within a matrix, or/and bordered by fragmented vegeta- tion, so strategies must be searched for improving and main- taining biodiversity within these areas.

Acknowledgments This work was supported by Fondo Mixto Consejo Nacional de Ciencia y Tecnología - Gobierno del Estado de Puebla (grant number PUE-2008-1-108571) to Ernesto I. Badano. The Administration of the State Park and Ecological Reserve Lázaro Cárdenas del Río “Flor del Bosque” granted permission to carry out the study and provided logistic support.

Authors’contributions This work is the result of the bachelor’s thesis of CHS. CHS, MC, GGJ, and EIB planned and designed experimental work. CHS, MNBL, FLC, MC, and GGJ executed experimental work. MNBL and FLC realized Lepidoptera collection and breeding. CHS and MC conducted data analyses and wrote the manuscript. CHS, MC, MNBL, GGJ, EIB, and FLC revised and made contributions to the manuscript.

References

Arizaga S, Martínez-Cruz J, Salcedo-Cabrales M, Bello-González MA (2009) Manual de encinos michoacanos. Secretaría de Medio Ambiente y Recursos Naturales, México, D.F., p 147

Badano E, Vergara C, García-Guzmán J, Barranco-Leó M de las N (2010) Encinos del Parque General Lázaro Cárdenas del Río “Flor del Bosque”: Technial Report. Gobierno del estado de Puebla, Puebla, 28 p

Baraza E, Gomez J, Hodar J, Zamora R (2004) Herbivory has a greater impact in shade tan in sun: response of Quercus pirenaica seedlings to multifactorial environmental variation. Can J Bot 82:357–364

Santin et al.

Barranco-León MN, Luna-Castellanos F, Vergara CH, Badadno EI (2016) Butterfly conservation within cities: a landscape scale approach inte- grating natural habitats and abandoned fields in central Mexico. Trop Conserv Sci 9(2):607–628

Basset Y (1991) The spatial distribution of herbivory mines and gall within an Autralian rain forest tree. Biotropica 23:271-281.

Bogaert B, Eysenrode SV, Impens I, Van Hecke P (2001) The interior-to- edge breakpoint distance as a guideline for nature conservation pol- icy. Environ Manag 27:493–500

Borror DJ, Triplehorn CA, Johnson NF (1992) An introduction to the study of insects. Harcourt Brace College Publishers, New York, p 800

Borror DJ, White RE (1970) Peterson field guides series: insects America north of Mexico. Houghton Mifflin, New York, p 416

Bossart J, Opuni-Frimpong E (2009) Distance from edge determines fruit-feeding butterfly community diversity in Afrotropical forest frag- ments. Environ Entomol 30:43–52

Buckley L (2000) The dynamics of temperate forest fragmentation: an examination of susceptibility to woody invasive species. Dissertation, Williams College, USA, 83

Connor E (2006) Effects of the light environment on oviposition prefer- ence and survival of a leaf’ mining moth, Cameraria hamadryadella (lepidóptera:Gracillariidae) on Quercus alba L. Ecol Entomol 31:179– 184

Costes-Quijano R, Meza AR, Macías-Juárez A, Berriel-Mastreta CA, Cortés-Atílano B, Martínez-Romero LE, Hernández-Bautista R (2006) Plan de manejo parque ecológico recreativo General Lázaro Cárdenas “Flor del Bosque”. Secretaría de Medio Ambiente y Recursos Naturales, Gobierno del Estado de Puebla, Puebla, 105

Crawley MJ (1985) Reduction of oak fecundity by low-density herbivore populations. Nature 314:163–164

Crawley MJ (1993) GLIM for ecologist. Blackwell Scientific Publications, Oxford, p 392

Cruz-Elizalde R, Ramírez-Bautista A, Wilson L, Hernández-Salinas U (2015) Effectiveness of protected areas in herpetofaunal conservation in Hidalgo, México. Herpetol J 25:41–48

De Carvalho Guimarães CD, Rodrigues Viana JP, Cornelissen T (2014) A meta-analysis of the effects of fragmentation on herbivorous insects. Environ Entomol 43:537–545

Extremera FM, Cobo A, Pérez-Rodríguez MC, Pérez Guerrero S, Vargas Osuna E (2004) El complejo de lepidópteros defoliadores de Quercus en la provincia de Córdoba. Boletín de Sanidad Vegetal Plagas 30: 203–209

Forkner RE, Marquis RJ, Lill JT, Corff JL (2008) Timing is everything? Phenological synchrony and population variability in leaf-chewing herbivores of Quercus. Ecol Entomol 33:276–285

Franklin JF, Spies TA, Van Pelt R, Carey AB, Thornburgh DA, Rae Berg D, Lindenmayer DB, Harmon ME, Keeton WS, Shaw DC, Bible K, Chen J (2002) Disturbances and structural development of natural forest ecosystems with silvicultural implications, using Douglas-fir forests as an example. Forest Ecol Manag 155:399–423

Hochwender C, Sork V, Marquis R (2003) Fitness consequences of her- bivory on Quercus alba. Am Midl Nat 150:246–253

Laurance WF (1991) Edge effects in tropical forest fragments- applications of a model for the design of nature-reserves. Biol Conserv 59:205–219

Laurance WF, Nascimento HEM, Laurance SG, Andrade A, Ewers RM, Harms KE, Luizão RCC, Rubeiro JE (2007) Habitat fragmentation, var- iable edge effects, and the landscape-divergence hypothesis. PLoS One 2:e1017. https://doi.org/10.1371/journal.pone.0001017

Laurance WF, Yensen E (1991) Predicting the impacts of edge effects in fragmented habitats. Biol Conserv 55:77–92

Liang Y, Stehlik I (2009) Relationship between shade and herbivory in Asarum canadense (wild ginger). Journal of Undergraduate Life Sciences 3:30–32

Lidicker WZJ (1999) Responses of mammals to habitat edges: an over- view. Landsc Ecol 14:333–343

Lincoln DE, Mooney HA (1984) Herbivory on Diplacus aurantiacus shrubs in sun and shade. Oecologia 64:173–176

López F (2004) Estructura y Función del borde. Ecosistemas 13:1–15 Maldonado-López Y, Cuevas-Reyes P, Oyama K (2016) Diversity of gall

wasps (Hymenoptera: Cynipidae) associated with oak trees (Fagaceae: Quercus) in a fragmented landscape in Mexico. APIS 10: 29–39

Marshall SA (2006) Insects their natural history and diversity. Firefly Books, New York, p 736

Medinaceli A, Miranda-Avilés F, Flores-Saldaña P, Gutierrez-Calucho E (2004) Herbivoría en relación al tamaño de la planta y a las diferencias de exposición de Pilea sp. (Urticaceae) en la Estación Biológica Tunquini, Cotapata, La Paz – Bolivia. Ecología en Bolivia 39: 4–8

Monreal JA, Salvador D, Mansilla J (1992) Contribución al conocimiento de los insectos defoliadores de la encina en la provincia de Albacete. Boletín de Sanidad Vegetal Plagas 18:395–405

Moore LV, Myers JH, Eng R (1988) Western tent caterpillars prefer the sunny side of the tree, but why? Oikos 51:321–326

Murakami M, Yoshida K, Hara H, Toda M (2005) Spatio-temporal varia- tion in lepidopteran larval assemblages associated with oak, Quercus crispula: the importance of leaf quality. Ecol Entomol 30:521–531

Murcia C (1995) Edge effects in fragmented forests: implications for conservation. TREE Trends Ecol Evol 10(2):58–62

Ochoa-Ochoa L, Urbina-Cardona JN, Vázquez L-B, Flores-Villela O, Bezaury-Creel J (2009) The effects of governmental protected areas and social initiatives for land protection on the conservation of Mexican amphibians. PLoS One 4(9):e6878. https://doi.org/10.1371/ journal.pone.0006878

Ordiales-Plaza R (1999-2000) Medición de objetos Versión 4.2. EEZA- CSIC (Estación Experimental de Zonas Áridas-Consejo Superior de Investigaciones Científicas)

Paint software. Versión 6.1, Microsoft ® Pereyra M, Pol RG, Galetto L (2015) Does edge effect and patch size

affect the interaction between ants and Croton lachnostachyus in fragmented landscapes of Chaco forest? APIS 9:175–186

Powell JA, Opler PA (2009) Moths of Western North America. University of California press, California, p 383

R Core Team (2015) R: a language and environment for statistical com- puting. R Foundation for Statistical Computing, Vienna, Austria URL http://www.R-project.org/

Ries L, Fletcher RJ Jr, Battin J, Sisk TD (2004) Ecological responses to habitat edges: mechanisms, models, and variability explained. Annu Rev Ecol Syst 35:491–522

Ries L, Sisk TA (2004) Predictive model of edge effects. Ecology 85:2917– 2926

Rodríguez-Correa H, Oyama K, MacGregor-Fors I, González-Rodríguez A (2015) How are oaks distributed in the Neotropics? A perspective from species turnover, areas of endemism, and climatic niches. Int J Plant Sci 176:222–231

Rossetti MR, González E, Salvo A, Valladares G (2014) Not all in the same boat: trends and mechanisms in herbivory responses to forest frag- mentation differ among insect guilds. APIS 8:593–603

Rzedowski GC de Rzedowski J (2001) Flora fanerogámica del Valle de México. Instituto de Ecología, A.C. y Comisión Nacional para el Conocimiento y Uso de la Biodiversidad, Michoacán, 1406

Santos BA, Benítez-Malvido J (2012) Insect herbivory and leaf disease in nature and human disturbed habitats: lessons from early-successional Heliconia herbs. Biotropica 44:53–62

Schowalter TD, Hargrove WW, Crossley DA Jr (1986) Herbivory in forest- ed ecosystems. Annu Rev Entomol 31:177–196

SEMARNAT-CONANP (2013–2018). Programa Nacional de Áreas Naturales Protegidas. Plan Nacional de Desarrollo. Gobierno de la República, México, 152

Eucalyptus Edge Effects on Quercus-Herbivore Interactions

Strauss S, Zangerl A (2002) Plant-insect interactions in terrestrial eco- systems. In: Herrera C, Pellmyr O (eds) Plant-animal interactions, an evolutionary approach. Blackwell Publishing, UK, pp 77–106

Summerville K, Boulware M, Veech J, Crist T (2003) Spatial variation in species diversity and composition. Conserv Biol 17:1045–1057

Vázquez ML, Valencia S, Nixon KC (2004) Notes on red oaks (Quercus sect. Lobatae) in eastern Mexico, with description of a new species, Quercus hirtifolia. Brittonia 56(2):136–142

Williams-Linera G, Dominguez-Gastelu V, Garcia-Zurita ME (1998) Microenvironment and floristics of different edges in a fragmented tropical rainforest. Conserv Biol 12:1091–1101

Williams-Linera G, Herrera F (2003) Folivory, herbivores, and environ- ment in the understory of a tropical montane cloud forest. Biotropica 35:67–73

Winter B (2014) A very basic tutorial for performing linear mixed effects analyses (tutorial 2). University of California, Merced, Cognitive and Information Sciences

Wirth R, Meyer ST, Leal IR, Tabarelli M (2008) Plant herbivore interac- tions at the forest edge. In: Lüttge U, Beyschlag W, Murata J. (eds) Progress in botany. Progress in botany, vol 69. Springer, Berlin, Heidelberg, pp 423–448

Publisher’s Note Springer Nature remains neutral with regard to ju- risdictional claims in published maps and institutional affiliations.

Santin et al.