**4. Morphometric analysis**

www.bee-bol.org).

350 The Dynamical Processes of Biodiversity – Case Studies of Evolution and Spatial Distribution

for Information Retrieval (www.tdwg.org/standards), which allows data to be exchanged

Currently one trend within the community of biodiversity informatics is to develop new standards for other contents, expanding from the current specimen/observation focus to other aspects of biodiversity, such as genomic data, interaction data (Saraiva et al., 2009) and species data, and multimedia data, such as images. The new contents will broaden the scope of the data networks and offer new possibilities for data analysis hopefully allowing address

Specimen and observation data are fundamental to develop distribution models using ecological niche concepts. Molecular data and images are key to identify a specimen and to

One of the biggest problems faced today by researcher is the lack of specialized personnel to identify biodiversity. In times of a rapidly changing world and fast loss of habitats and biodiversity, it is almost impossible to measure the existing species in the ecosystems. There is a lack of identification keys and genera revisions. The taxonomists are few and normally have more complexes problems to focus the simple species identification for the general public or to scientists from other science areas. The development of alternative tools to

At this point, the astonishing development of molecular biology in the last years has indicated new alternatives that can be used to identify species. Since the final of the last century, mitochondrial DNA (mtDNA) has been used as a very interesting and powerful alternative tool. This molecule has an enormous potential due to extremely peculiar and unique characteristics, like being a small circular genome, with high evolutionary rates but well conserved in animals (Arias et al., 2003). In neotropical stingless bees, it has been largely used in populational studies and in the evaluation of genetic diversity (Francisco et

Although very controversial, the use of this molecule to identify species was proposed in 2003. Based on the principle that differences on the sequence of the genes are greater between species than within species, the proposition consists in sequencing approximately 650 base pair from the beginning of the Cytochrome Oxidase I (CO-I) gene and comparing it among the species (Herbert et al., 2003). According to recent revisions, the studies show the efficiency of these genic regions to discriminate species and it is working well in the vast majority of the animal cases studied until that moment (more than 95%) (Vogler and

In fact, the use of a mtDNA sequence to identify cryptic species is constantly reported in literature. However, the standardization of the procedures and the establishment of some guidelines are the novelty in the DNA barcode proposition (Brown et al., 1999; Mitchell and

the possibility of an operation to be largely used, since the sequencing methodologies

the obligatory deposition of vouchers in entomological collections, in order to facilitate

Samways, 2005). Briefly, the following sequence is proposed in this procedure:

future studies of combined morphological + molecular studies;

among different systems, using agreed upon standards, such as DwC.

study the relationship between individuals and populations.

assess biodiversity other than traditional taxonomy is an urgent need.

al., 2001; 2008; Brito & Arias, 2005; May-Itzá et al., 2010).

Monaghan, 2007; Waugh, 2007).

the standardization of the region to be used;

are becoming more and more accessible;

**3. DNA barcode** 

issues that are even closer to societal needs, cross-cutting different disciplines.

The first attempts to classify bee subspecies of *Apis mellifera* was based mainly in differences in color and body size. However, since there is a great superimposition in these parameters, most of the classification systems based in these characteristics failed in correctly identify the individuals (Ruttner, 1988). In 1940, Goetze proposed a large number of measures, to be taken from several parts of the bee body in order to better differentiate the geographical ecotypes present all over its wide geographical distribution that encompasses Africa, Europe and parts of Asia.

However, all the analysis used until the moment were based in uni-variate statistics, which takes into consideration only one measure at a time and the range of the measures often overlap and turn more difficult to achieve a precise identification. It was only after the works of DuPraw (1964, 1965a; b) that the usage of multivariate statistics was proposed and, with help of Principal Component Analysis and Discriminant Analysis that the identifications became more precise. An important advance is also proposed in this series of works, where DuPraw (1965a) indicated the use of measures that are independent of size, like angles between vein junctions in the wings, avoiding the environmental effects, like food availability, parasites and others.

After these propositions and a series of small studies, it is published a guide to discriminate the subspecies of *A. mellifera* (Ruttner et al., 1978). In this work, the authors propose approximately 40 measures to be taken from several parts of at least 20 bees per colony to achieve a good confidence in the classification. It was based on morphometric results that the existence of evolutionary branches in *A. mellifera* (Ruttner et al., 1978) that were later confirmed by mitochondrial DNA (Franck et al., 2000), microsatellites (Estoup et al., 1995) and SNPs (Withfield et al., 2006). In spite of being very informative and confident, this kind of analysis is often very time consuming.

More recently, allied to the development of computational methods, the analysis became faster and some of them completely automated as ABIS (Automated Bee Identification

Biodiversity in a Rapidly Changing World: How to Manage and Use Information? 353

(Rattanawannee et al., 2010), and also the efficiency in discriminating the species when large datasets are used at the same time, as the correct identification of 93% of the individuals into the respective group for 34 different species of stingless bees (Francoy et al., unpublished

Together with the efficiency in identifying species based only in the patterns of wing venation, perhaps one of the most important applications of this kind of analysis is the possibility of mapping intra-specific variability within a species and consequently, tracking the geographical origin of samples. It allows the researchers to evaluate this variability in several geographic scales. The stingless bee *Melipona beecheii* has a geographical distribution that ranges from Mexico to Costa Rica, where it inhabits the most varied environments. An analysis of the patterns of wing venation of bees from Mexico, Nicaragua, El Salvador, Guatemala and Costa Rica indicated marked differences among the populations, correctly re-assigning 87% of the individuals to the respective group (Francoy et al., 2011). It was also demonstrated that it is valid for other examples, like bees from the genus *Peponapis* in North America (Bischoff et al., 2009), *Apis florea* (Kandemir et al., 2009) and *Apis mellifera*  populations (Özkhan and Kandemir, 2010) and *Nannotrigona testaceicornis* sub-populations

Another geometric approach that is very promising is the outline of wing cells. It has already been demonstrated that features extracted from a single wing cell can discriminate *Apis mellifera* subspecies (Francoy et al., 2006)*.* Based on this principle, it was proposed that people already knew how to manage colonies to transportation around 3000 years ago. In an archaeological excavation in the middle of the Jordan valley in Northern Israel it was found what appears to be a well-organized apiary. Two of the hives contained charred honey comb remains with many honey bee body remains. Although most remains were damaged, two wings with clear cells were of sufficient quality to perform morphometric measurements comparable to those available for present-day subspecies over the entire distribution range of *A. mellifera*. However, as only small parts were available, only feature extracted from single wing cells were compared. It was determined that the wings belonged to the subspecies currently living in parts of Turkey, instead of the one living in Israel. Since the climatic data indicated no extreme climate and vegetation change in the last 3000 years, the authors concluded that the beekeepers already knew how to transport colonies across long distances and kept importing bees from Turkey because of the more suitable behaviour

of the *A. m. anatoliaca* bees rather than the original *A. m. siriaca* (Bloch et al., 2010).

Despite effective species discrimination from application of landmark or outline-based methods used independently, the combined results of these two methods is only now being investigated. In an exploratory study, five species of *Euglossa* were analyzed using landmark and outline based methods in order to compare the efficiency of both. Regarding the landmark analysis, 18 landmarks were used and achieved 84% of correct identifications. In the outline based analysis, a complete exploratory characterization of all wing cells was made and the wing cells that better discriminated the five species correctly re-assigned 77% of the individuals to the respective group. However, when using the features extracted in both analysis in a combined matrix, the correct classification rates achieved 91% (Francoy et

In order to improve the process and to make the analysis faster and more precise, new tools are being developed for a complete automation of the system like algorithms to automatic identification of the landmarks (Bueno, 2010) and new processes of features extraction that make the entire process more reliable and efficient (Buani, 2010). The automated

results).

(Mendes et al., 2007).

al., unpublished results).

System) (Schröder et al., 2001). The first step in this process was the construction of a semiautomated system based on features extracted from the images of the forewings, in which the user had to plot landmarks in the wing vein intersections (Schröder et al., 1995). A full automated version of the software was developed with some modifications, like the automated identification of the landmarks and the implementation of a non-linear discriminant analysis, which improved the identification rates of the individuals (Schröder et al., 2001). In this process, ABIS extract more than 300 features related to create a "fingerprint" of each species. This pattern is stored in a databank and each new wing loaded in the software is compared to the databank in order to identify the species. It was very efficient in discriminating the European species of the genus *Andrena, Colletes* and *Bombus* (Schröder et al., 2001), Africanized honey bees (Drauschke et al., 2007; Francoy et al., 2008) and also *Euglossa* species (Francoy et al., unpublished data).

Another morphometric technique that is presenting very interesting results concerning shape variation is geometric morphometrics (Bookstein, 1991). While studies with standard morphometrics analyze shape variation using co-variation of pairs of linear measures, geometric morphometrics is based on the variation of the relative positions of the landmarks and therefore, is able to describe more clearly any changes in shape and also to graphically reconstruct these differences (for a detailed description of the method, see Rohlf and Marcus, 1993).

The first attempt to use this methodology in the patterns of wing venation to differentiate bee groups was done in Africanized honey bees and in the subspecies that formed this hybrid. The relative warps analysis of the landmark positions in the wing was able to correctly classify 85% of the individuals in the correct group and the higher error rates were found in two subspecies that belong to the same evolutionary branch (Francoy et al., 2008). It is important to state that these bees are not easily distinguishable even for the well established standard morphometric methods. Another interesting result from this work is the possibility of correctly identify 99.2% of the Africanized sample, which is always a concern in areas newly occupied for these bees. Additionally, the usage of these methods allows a quicker identification than the traditional methods, once it can be done in a few minutes while the identification through standard morphometrics is more time consuming, around a few hours per colony. Still in honey bees, it was already demonstrated that other European subspecies (Tofilsky, 2008) can be identified using this methodology as well as different species from the genus *Apis* (Rattanawannee et al., 2010).

In stingless bees, the first work using this methodology demonstrated the power of the technique to identify cryptic biodiversity (Francisco et al., 2008). Colonies from two distinct populations of *Plebeia remota* kept in the same apiary for more than 10 years do not presented any gene flow. Until that moment, the populations were considered as the same species. This result was reinforced by other molecular markers, like mitochondrial DNA and cuticular hydrocarbons, which pointed in the same direction of the morphological data. It was also very informative to discriminate species with very little or no external morphological differences, like species from the genus *Eubazus* (Villemant et al., 2007), *Bombus* (Aytekin et al., 2007) and *Euglossa* (Francoy et al., unpublished results). Other works also indicated the efficiency of the technique in stingless bees. When working with bees from the same genus, studies showed 93.4% of success in the discrimination of 6 species of *Plebeia* (Francoy et al., unpublished results)*.* It was also demonstrated the discrimination of sub-populations of *Nannotrigona testaceicornis* (Mendes et al., 2007), differences between the wings of males and workers in stingless bees (Francoy et al., 2009) and in honey bees

System) (Schröder et al., 2001). The first step in this process was the construction of a semiautomated system based on features extracted from the images of the forewings, in which the user had to plot landmarks in the wing vein intersections (Schröder et al., 1995). A full automated version of the software was developed with some modifications, like the automated identification of the landmarks and the implementation of a non-linear discriminant analysis, which improved the identification rates of the individuals (Schröder et al., 2001). In this process, ABIS extract more than 300 features related to create a "fingerprint" of each species. This pattern is stored in a databank and each new wing loaded in the software is compared to the databank in order to identify the species. It was very efficient in discriminating the European species of the genus *Andrena, Colletes* and *Bombus* (Schröder et al., 2001), Africanized honey bees (Drauschke et al., 2007; Francoy et al., 2008)

Another morphometric technique that is presenting very interesting results concerning shape variation is geometric morphometrics (Bookstein, 1991). While studies with standard morphometrics analyze shape variation using co-variation of pairs of linear measures, geometric morphometrics is based on the variation of the relative positions of the landmarks and therefore, is able to describe more clearly any changes in shape and also to graphically reconstruct these differences (for a detailed description of the method, see Rohlf and

The first attempt to use this methodology in the patterns of wing venation to differentiate bee groups was done in Africanized honey bees and in the subspecies that formed this hybrid. The relative warps analysis of the landmark positions in the wing was able to correctly classify 85% of the individuals in the correct group and the higher error rates were found in two subspecies that belong to the same evolutionary branch (Francoy et al., 2008). It is important to state that these bees are not easily distinguishable even for the well established standard morphometric methods. Another interesting result from this work is the possibility of correctly identify 99.2% of the Africanized sample, which is always a concern in areas newly occupied for these bees. Additionally, the usage of these methods allows a quicker identification than the traditional methods, once it can be done in a few minutes while the identification through standard morphometrics is more time consuming, around a few hours per colony. Still in honey bees, it was already demonstrated that other European subspecies (Tofilsky, 2008) can be identified using this methodology as well as

In stingless bees, the first work using this methodology demonstrated the power of the technique to identify cryptic biodiversity (Francisco et al., 2008). Colonies from two distinct populations of *Plebeia remota* kept in the same apiary for more than 10 years do not presented any gene flow. Until that moment, the populations were considered as the same species. This result was reinforced by other molecular markers, like mitochondrial DNA and cuticular hydrocarbons, which pointed in the same direction of the morphological data. It was also very informative to discriminate species with very little or no external morphological differences, like species from the genus *Eubazus* (Villemant et al., 2007), *Bombus* (Aytekin et al., 2007) and *Euglossa* (Francoy et al., unpublished results). Other works also indicated the efficiency of the technique in stingless bees. When working with bees from the same genus, studies showed 93.4% of success in the discrimination of 6 species of *Plebeia* (Francoy et al., unpublished results)*.* It was also demonstrated the discrimination of sub-populations of *Nannotrigona testaceicornis* (Mendes et al., 2007), differences between the wings of males and workers in stingless bees (Francoy et al., 2009) and in honey bees

and also *Euglossa* species (Francoy et al., unpublished data).

different species from the genus *Apis* (Rattanawannee et al., 2010).

Marcus, 1993).

(Rattanawannee et al., 2010), and also the efficiency in discriminating the species when large datasets are used at the same time, as the correct identification of 93% of the individuals into the respective group for 34 different species of stingless bees (Francoy et al., unpublished results).

Together with the efficiency in identifying species based only in the patterns of wing venation, perhaps one of the most important applications of this kind of analysis is the possibility of mapping intra-specific variability within a species and consequently, tracking the geographical origin of samples. It allows the researchers to evaluate this variability in several geographic scales. The stingless bee *Melipona beecheii* has a geographical distribution that ranges from Mexico to Costa Rica, where it inhabits the most varied environments. An analysis of the patterns of wing venation of bees from Mexico, Nicaragua, El Salvador, Guatemala and Costa Rica indicated marked differences among the populations, correctly re-assigning 87% of the individuals to the respective group (Francoy et al., 2011). It was also demonstrated that it is valid for other examples, like bees from the genus *Peponapis* in North America (Bischoff et al., 2009), *Apis florea* (Kandemir et al., 2009) and *Apis mellifera*  populations (Özkhan and Kandemir, 2010) and *Nannotrigona testaceicornis* sub-populations (Mendes et al., 2007).

Another geometric approach that is very promising is the outline of wing cells. It has already been demonstrated that features extracted from a single wing cell can discriminate *Apis mellifera* subspecies (Francoy et al., 2006)*.* Based on this principle, it was proposed that people already knew how to manage colonies to transportation around 3000 years ago. In an archaeological excavation in the middle of the Jordan valley in Northern Israel it was found what appears to be a well-organized apiary. Two of the hives contained charred honey comb remains with many honey bee body remains. Although most remains were damaged, two wings with clear cells were of sufficient quality to perform morphometric measurements comparable to those available for present-day subspecies over the entire distribution range of *A. mellifera*. However, as only small parts were available, only feature extracted from single wing cells were compared. It was determined that the wings belonged to the subspecies currently living in parts of Turkey, instead of the one living in Israel. Since the climatic data indicated no extreme climate and vegetation change in the last 3000 years, the authors concluded that the beekeepers already knew how to transport colonies across long distances and kept importing bees from Turkey because of the more suitable behaviour of the *A. m. anatoliaca* bees rather than the original *A. m. siriaca* (Bloch et al., 2010).

Despite effective species discrimination from application of landmark or outline-based methods used independently, the combined results of these two methods is only now being investigated. In an exploratory study, five species of *Euglossa* were analyzed using landmark and outline based methods in order to compare the efficiency of both. Regarding the landmark analysis, 18 landmarks were used and achieved 84% of correct identifications. In the outline based analysis, a complete exploratory characterization of all wing cells was made and the wing cells that better discriminated the five species correctly re-assigned 77% of the individuals to the respective group. However, when using the features extracted in both analysis in a combined matrix, the correct classification rates achieved 91% (Francoy et al., unpublished results).

In order to improve the process and to make the analysis faster and more precise, new tools are being developed for a complete automation of the system like algorithms to automatic identification of the landmarks (Bueno, 2010) and new processes of features extraction that make the entire process more reliable and efficient (Buani, 2010). The automated

Biodiversity in a Rapidly Changing World: How to Manage and Use Information? 355

composition at the whole range of their host plant distribution, Espíndola et al. (2011) found geographically structured variability of the prevailing visitor. They suggested that climate is driving the specificity of this interaction, by potentially affecting the phenology of one or both interacting species, providing an example of the direct effect that the abiotic

This is in accordance with Thompson (2005) who suggested a geographic mosaic theory of coevolution stating that interspecific interactions commonly exhibit geographic selection mosaics and trait remixing among populations. From this view, the form and trajectory of coevolutionary selection vary across landscapes. In addition, gene flow and metapopulation dynamics continually shift traits among populations, thereby continually altering the

Laine (2009) reviewed 29 studies that support this theory, concluding that natural coevolutionary selection produces genetic differentiation among populations and may be an important mechanism promoting diversity in nature given how different types of interactions show divergence, and how variable the causes promoting such divergence are. One of the remarkable results of this review is the spatial scale over which it is possible to find divergent coevolutionary trajectories. Variation was detected in populations separated by some hundreds of kilometers highlighting the potential for the environment to create geographically variable selection trajectories. For example, analyzing a rare and endangered solitary bee (*Colletes floralis*), Davis et al. (2010) found an extremely high genetic differentiation among populations at the extreme edges of the species range. Also, Pellissier et al. (2010) analyzed how the traits of different pollination syndromes influence the distributions of plant species in interaction with pollinators. They used a combination of environmental descriptors and found a potential effect of the pollinator on the spatial distribution of plant species. Also, analyses of a system involving the Japanese camellia and its obligate seed predator, found that the sizes of the plant defensive apparatus (pericarp thickness) and the weevil offensive apparatus (rostrum length) clearly correlated with each

Therefore, intermingled with environmental (abiotic) and interaction (biotic) features, geographical distribution is also related to species evolutionary trends, determining patterns of genetic diversity and trait variation across space. New approaches are necessary to analyze the importance of these complex features. Recently, Pavoine et al. (2011) suggested a framework based on a mathematical method of ordination to analyze phylogeny, traits, abiotic variables and space in a plant community. Another example can be found in Diniz Filho et al. (2009) proposing an integrated framework to study spatial patterns in genetic diversity within local populations, coupling genetic data, SDM and landscape genetics. Also, Kuparinen and Schurr (2007) developed a framework to link the spatio-temporal dynamics of plant populations and genotypes, and a similar approach was suggested for modeling the variation of geographical distribution of animals in a climate change scenario

To analyze the multiple drivers shaping the species geographical distribution is a challenge that will be met by integrating different fields of research. In order to attain the objective of predicting impacts on species distribution due to global changes, it is necessary to consider that species are genetically heterogeneous entities and, in order to protect them, it is necessary to protect its genetic diversity. As species diversity might act as insurance against environmental changes, genetic diversity should also have the potential to protect communities from environmental variability (Lavergne et al., 2010). Taubmann et al. (2011)

other across geographically structured populations (Toju and Sota, 2006).

environment can have on the plant–insect interaction.

structure of local selection.

(Kearney and Porter, 2009).

identification system uses two computational algorithms to complete the recognition and classification of bee species. The first algorithm, named kNN (k-Nearest Neighbour), is used to select and extract morphometric features related to the distances between the landmarks plotted in the wing veins intersections from the pictures. The second algorithm, named FkNN (Fuzzy kNN) implements a variation of the Fuzzy Logic for species classification. For an optimized result, the features selection involves a statistical analysis which evaluates the better landmarks for the classification process and only the most informative are used in the species characterization in the Fuzzy Logic (Buani, 2010).

The morphometric analysis of forewing is a very powerful tool to describe species variation and also to identify species based on landmark and outline morphometric methods. Allied to that, morphometric analysis is a fast, inexpensive and informative method to be used in the characterization of species and its variation.
