Aspilota-group (Hymenoptera: Braconidae: Alysiinae) diversity in Mediterranean Natural Parks of Spain

Abstract This work analyses the biodiversity of the Aspilota-group (Hymenoptera: Braconidae: Alysiinae) in three Mediterranean Natural parks: Natural Park of La Font Roja, Natural Park of Las Lagunas de la Mata-Torrevieja and Natural Park of La Tinença de Benifassà. Samples were carried out from April 2004 to December 2007. In total, 822 specimens, belonging to 52 species, were collected. Alpha, beta and gamma diversities were analysed, and the Tinença Park was proven to have higher diversity than the Font Roja and Torrevieja. Also, the structure of the Aspilota-group community was analysed.


Introduction
Mediterranean ecosystems are very important in terms of biodiversity, and are thus considered hotspot areas (Myers et al. 2000). Landscapes and habitats grow in complexity over time, as a consequence of ecological processes. For example, Mediterranean forest landscapes rich in evergreen species frequently intersect with brushwood, pasture, farming and ranching areas. In close proximity to these areas, however, it is often possible to identify zones that have been reclaimed by highly diverse natural communities after the cessation of human intervention. Despite the huge resistance displayed by Mediterranean biotopes to human pressure, isolation and fragmentation are unavoidable (Pungetti 2003), resulting in the emergence of isolated patches in the landscape.
In land environments, the information provided by arthropods can be very valuable for the adoption of measures aimed at guaranteeing the diversity and welfare of protected forests (Pyle et al. 1981, Pearson and Cassola 1992, Kremen et al. 1993, especially insects with a high sensitivity to alterations in environmental resources and conditions. Parasitoids Hymenoptera of the Braconidae family, with around 40,000 catalogued species, are especially pertinent in this respect due to their particular biology (Wharton et al. 1997).
Braconidae are the second largest family within Hymenoptera; the majority of species are primary parasitoids of immature stages of Lepidoptera, Coleoptera and Diptera (Sharkey 1993). These wasps are of enormous ecological interest because of their role in controlling the population of phytophagous insects, causing direct effects in the host species' population size and indirect effects on the diversity and survival of host plants (LaSalle and Gauld 1992). Additionally, they can indicate the presence or absence of said populations (Matthews 1974, LaSalle andGauld 1992). Finally, some species can also be relevant from an economic point of view, because of their potential for pest control (González and Ruíz 2000).
Because of the type of relationship established between Braconidae populations and host species, and the effect that climatic factors and human activity pose upon this, we can consider that Braconidae (especially those adopting koinobiont strategies) are a valid parameter for the determination of human effects on these communities and the assessment of specific diversity within a region (González and Ruíz 2000).
average temperatures throughout the year (15-20°C), and the low average rainfall, the park is classified is dry and thermo-Mediterranean.
The Natural Park of Las Lagunas de la Mata-Torrevieja is located to the south of Alicante province, and extends over 3,700 ha, 2,100 of which are covered by water. The park is notable for its saline soils, extensive wild orchid population (Orchis collina Banks and Sol. ex Russell), differentiated areas of Senecio auricula Bourgeau ex Coss and salt marsh plants of the genus Limonium, reed and bulrush areas with abundant grass plants such as Arthrocnemum sp. and Juncus sp., and Mediterranean areas populated by Quercus coccifera L., Pinus halepensis Mill. and Thymus sp. The climate is arid with an annual rainfall below 300 mm and high temperatures.
The Natural Park of La Tinença de Benifassà is located to the north of Castellón province, and extends over approximately 25,814 ha. The park covers an extensive and wellpreserved mountainous area, encompassing numerous and widely varied landscapes associated with medium and high-altitude Mediterranean regimes and hosting a high biodiversity of fauna and flora. It is possible to differentiate forests of Pinus sylvestris L., Pinus uncinata Mill. and Fagus sylvatica L., Juniperus communis L., and Quercus ilex L., etc., alternating with crops of Prunus sp., Corylus sp., etc. Climate conditions are continental humid, with annual average temperatures below 12°C: freezing conditions are possible throughout most of the year. Rainfall varies in different zones according to topographical features, and the annual precipitation ranges from 600 to 1,000 l/m . The park is contained within the supramediterranean bioclimate.

Sampling Design and Data Collection
Sampling stage covered the period among April 2004 and December 2007. During this period, in each natural park, a Malaise trap to collected specimens was placed. Weekly, each area was visited to replace the collecting bottle. Specimens captured were preserved in 70% ethanol until final preparation.
Once separated, the specimens were determined by subfamily keys of van Achterberg (1990) to work only with Alysiinae specimens. Subsequently, the identification to genera was carried out using key. Finally, species identification was did it by Fischer (1993b), Fischer (2003), Fischer et al. (2008a), Fischer et al. (2008b) and Tobias' keys (Tobias et al. 1986). The studied specimens are deposited with a bar code labels in the Entomological Collection at the University of Valencia (Valencia, Spain; ENV). General distribution data were provide from Yu et al. (2012). 2

Data analysis
Once the specimens of Aspilota-group had been identified, alpha, beta and gamma biodiversity indexes for each trap and habitat were calculated to gain insight into the richness, abundance, dominance and complementarity values of each area.
Alpha diversity reflects the richness in species of a homogeneous community. This sort of diversity was measured by taxa richness, abundance and dominance.

•
Taxa richness: used for valuing richness of sampling areas. It was measured using the Margalef index, a measure of specific richness that transforms the number of species per sample into the proportion to which the species are added by expansion of the sample, establishing a functional relationship between number of species and total number of specimens (Moreno 2001). • Species richness estimators: It was measured using Chao 2 to know what percentage of the total known species of possible species (Moreno 2001). • Abundance: used for valuing faunal composition of a given area (Magurran 1988). This was undertaken using the Shannon-Weaver index because it measures equity, indicating the degree of uniformity in species representation (in order of abundance) while considering all samples. This index measures the average degree of uncertainty that predicts which species an individual randomly picked from a sample belongs to (Magurran 1988, Moreno 2001, Villarreal et al. 2004). • Dominance: occurrence of genera or dominance value was calculated with the Simpson index, often used to measure species dominance values in a given community, with negative values thus representing equity. It measures the representativity of the most important species without considering the other species present. It expresses the probability that two individuals randomly picked from a sample will belong to the same species (Magurran 1988). • Community structure: In order to complement the diversity analyses and enquire into community structure, log-series, log-normal and broken-stick models were also applied (Magurran 1988). The log-series model represents a community composed of a few abundant species and a high number of rare species. The broken-stick model refers to maximum occupation of an environment with equitable sharing of resources between species. Finally, the log-normal reflects an intermediate situation between the two (Soares et al. 2010). Using the data obtained from the parks, each of these models was applied to calculate the expected number of species, and log grouping species according to abundance (Magurran 1988, Tokeshi 1993, Krebs 1999. To test the significance of the models, the expected species values were compared with those of the observed species through chisquare analysis (Zar 1999).
Beta diversity is the degree of change or substitution in species composition between different communities within the same landscape. In order to measure beta diversity, Jaccard and Complementarity indexes were used and cluster analyses were also performed.
• Jaccard index: relates the total amount of shared species to the total amount of exclusive species. It is a qualitative coefficient, the interval of which will go from 0 when no species are shared between both sites to 1 when both sites have an identical composition (Moreno 2001, Villarreal et al. 2004). • Complementarity index: indicates the degree of similarity in species composition and abundance between two or more communities (Moreno 2001, Villarreal et al. 2004). • Cluster analysis: employed to calculate the degree of correlation based on similarity/dissimilarity. For the calculation of these values, statistics-processing software PAST was used (Hammer et al. 2001).
Finally, gamma diversity measurement indicates the diversity value of all environments under study, as expressed in the richness indexes for each area (alpha diversity) and the difference between them (beta diversity) (Schluter andRicklefs 1993, Villarreal et al. 2004).
However, the species were not evenly distributed when different Natural Parks are considered separately. Thus, 39 species were identified in the Natural Park of La Tinença de Benifassà (Tinença), 23 were identified in the Natural Park of Carrascal de La Font Roja (Font Roja) and 21 were identified in the Natural Park of Las Lagunas de la Mata-Torrevieja (Torrevieja).
The genus Dinotrema was the most abundant with 343 examples, followed by the genera Synaldis (271) and Aspilota (108). On the other hand, when analysing the number of captures, it was observed that 383 individuals were collected in Tinença, 257 in Torrevieja and 182 in Font Roja. In Tinença the most captured genera was Dinotrema with 202 specimens followed by Synaldis with 95. However, in Torrevieja and Font Roja the most captured genera was Synaldis with 105 and 71 specimens respectively, followed by Dinotrema with 93 and 48.
The Margalef index (D ) shows that Natural Park of Tinença hosted a higher species richness with D = 6.389, while Font Roja reached a value of 4.228 and Torrevieja 3.604. These values might be so discordant as a consequence of the identified species differing widely from Tinença (39 species) to other habitats. Font Roja and Torrevieja has a similar D value because its species number is very close (23 and 21 species respectively).
On the other hand, with the estimators of species richness (Chao 2), it is possible conclude that the Natural Park where our sampling effort has enabled a closer approximation to the estimated maximum richness is Font Roja with a value of 94.62%, followed by Tinença and Torrevieja with values of 82.97% and 79.75% respectively.
When analyzing the structure of the community it is needed to distinguish between two types of analysis: proportional abundance indices or parametric models.
First, the community structure is studied by proportional abundance indices in which differentiate dominance indices as Simpson or Berger-Parker and equity index as Shannon-Wiener.
The results obtained with the Simpson and Berger-Parker index (Table 1) show a dominance of the community structure by one or more species with high population abundance. The Shannon index suggested a similar trend in the distribution of dominant genera; discrepancies were merely due to different numbers of rare genera (those represented by few specimens). Finally, applying parametric models, ( Table 2) the analysis of the Aspilota-group community structure showed that Font Roja and Torrevieja present compliance with the log-series model indicating that these communities have an unstable structure, composed by few abundant species and a large number of rare species. These results show that habitat does not determine community structure because the sampling area presents very specific botanical and faunal composition and climatic conditions.

Mg
Mg Mg Table 1.
Expected frequency of species (exp f) according to abundance models (log-series, log-normal and broken-stick) for the Aspilota-group community.  However, Tinença shows compliance with log-series and log-normal models presenting, more or less, the same p-value (0.501 and 0.513 respectively). This fact could be indicating two types of behaviour. On the one hand, this community could be unstable and composed by few abundant species and large number of rare species. And, on the other hand, it could be indicating that the specimens number of this community is conditioned by a large number of factors associated with high temperatures and low rainfall that occur in this area causing that species must adapt to very strict conditions.
In order to obtain beta diversity (similarity/dissimilarity) values between the different areas under consideration, the Jaccard index was calculated. The resulting value indicated a certain degree of dissimilarity between the Natural Parks although, Font Roja and Tinença are the closest parks (I = 0.377) while Font Roja and Torrevieja are the farthest parks (I = 0.189). These results were also observed in the Jaccard cluster obtained through cluster analysis, of which the level of correlation was r = 0.8863 (Fig. 1).
Expected frequency of species (exp f) according to abundance models (log-series, log-normal and broken-stick) for the Aspilota-group community.
However, applying the Principal Component Analysis (PCA) (Fig. 2) shows that there are many unique species to each Natural Park (16 for Tinença, 7 for Torrevieja and 6 for Font Roja) while the rest of species are usually present shared (17 for Font Roja-Tinença, 12 for Tinença-Torrevieja and 7 for Font Roja-Torrevieja). This could be due to the fact that Tinença and Font Roja are Mediterranean forests while Torrevieja is a lagoon.
The indices of species replacement by the Whittaker index (Table 3) show that the Natural Park of La Tinença de Benifassà has not a lot of replacement with species from other Natural Parks, while, Torrevieja and Font Roja show some replacement. This relationship could be possible thanks that these natural parks are close while Tinença is far.  The Complementarity index (C) suggested that the Font Roja and Torrevieja has the highest complementarity (0.810) followed by Tinença and Torrevieja with 0.723 and Tinença and Font Roja with 0.622. These results showed a fair degree of complementarity, but also indicated the presence of different species in each habitat (Table 3). This fact could be explained because these natural parks are close to each other while Tinença is farther apart.
Finally, gamma diversity reached a value of 52.954, which is practically identical to the value of the total species richness caught in the three Natural Parks (species number = 53).

Discussion
Regarding the faunistic study, four species captured are new records for Spain: Aspilota delicata, Aspilota procreata, Dinotrema costulatum and Dinotrema crassicostum. While, regarding the biodiversity study, it is possible to see that the Natural Park of La Tinença de Benifassà presents greater abundance and species diversity, followed by the Font Roja and Torrevieja parks. On the other hand, the analysis of the structure of the network has showed that the Font Roja and the Torrevieja Natural Parks show a model of community that matches the log-series model. This indicates that these communities are unstable and are composed of few abundant species and a large number of rare species. While the Table 3.
Whittaker index and Complementarity index values for Aspilota-group between Natural Parks. community of the Aspilota-group present in the Tinença, is adapted to the models of logseries and log-normal. This demonstrates that the structure of the community is not determined by the habitat, but conditioned by a large number of factors associated with the high temperatures and low rates of precipitation, which may force the species to adapt to strict environmental conditions. Furthermore, when comparing parks, it can be seen that La Tinença and La Font Roja show the most similarities between each other, whilst the Font Roja Park and the Torrevieja Park show a larger group of species that complement each other.
On the other hand, checking with the studies realized in other areas of Spain as Artikutza about the Aspilota-group show that this group was the most abundant captured with approximately 75.77% (Peris-Felipo and Jiménez-Peydró 2011). The information about the abundance is very interesting due to the relationships that these parasitic wasps have with their hosts. This information could be used to estimate the biodiversity appearing in each area.
Finally, we conclude that, although this study was conducted to determine the diversity and community structure of the Aspilota-group, it is recommended further studies of Braconidae in different areas together with DNA-barcode studies to increase the knowledge of this large group that still remains largely unknown.