Characters in Thysanoptera Mes 7(2) 16-37

Embed Size (px)

Citation preview

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    1/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2): 16ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    AutorCorrespondencia: [email protected]*

    Revisionary paper

    Characters in Thysanoptera

    AXEL P. RETANA-SALAZAR* Programa Universitario de Biologa Aplicada (PUA), Centro de Investigacin en Estructuras Microscpicas(CIEMic), Ciudad de la Investigacin, Universidad de Costa Rica, 2060.

    R ESUMEN : Thysanoptera es un grupo de alta homoplasia como ha sido establecido en la literatura pormltiples autores. Esto ha tenido un efecto en la discusin entre los taxnomos que invierten grandescantidades de tiempo en la discusin de la validez de los caracteres utilizados en este grupo de insectos. Esto

    se refleja en mltiples trabajos de taxonoma en los que se describen especies y grupos sin unafundamentacin filogentica. De igual forma decisiones tomadas en funcin de resultados filogenticos sonmuchas veces cuestionados ya que los mismos van en contra de una clasificacin ms cmoda. En estetrabajo se presenta una revisin corta de las tendencias y problemas en los diferentes tipos de caracteres quese utilizan en Thysanoptera en la actualidad y la necesidad de tener mayor cuidado con el trato de algunosde estos caracteres en la toma de decisiones taxonmicas, en las cuales se proponen cambios en laclasificacin biolgica.

    PALABRAS CLAVE : Filogenia, caracteres moleculares, caracteres morfolgicos, caracteres biolgicos,caracteres teratolgicos, cladstica., Thysanoptera, Thripidae.

    ABSTRACT : Thysanoptera is a group of high homoplasy as has been established in the literature by many

    authors. This has had an effect on the debate among taxonomists investing large amounts of time discussingthe validity of the characters used in this group of insects. This is reflected in many taxonomic works,described species and phylogenetic groups without a foundation. Similarly decisions based on phylogeneticresults are often questioned since they go against a more comfortable classification. This paper presents ashort review of trends and issues in different types of characters used in Thysanoptera today and the needfor a more careful treatment of some of these characters in taxonomic decisions in which changes in the

    biological classification are proposed.

    K EY WORDS : Phylogeny, molecular characters, morphological characters, biological characters, teratologicalcharacters, cladistics., Thysanoptera, Thripidae.

    I NTRODUCTION

    Thysanoptera is a complex group especially because of its high level of homoplasy (Gauld &Mound 1982, Mound & Palmer 1983, Retana-Salazar 1998, 2000). However it is one of thegroups with fewer amounts of formal

    phylogenetic analysis and with a very common practice of -taxonomy of the specialists that

    are more occupied in the description of new

    species than in the determination of the naturalgroups.

    The recognized authorities focused in the -taxonomy producing several hundreds of pages

    per year full of subjective criteria about theimportance of the characters used by eachauthor. This practice was abandoned in the

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    2/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    17

    majority of the groups, where several authoritiesdont consider the publication of speciesdescription in absence of good phylogeneticanalysis and revisionary works as good exercise(Retana-Salazar 1998, 2000, 2006, Retana-Salazar & Ramrez Morales 2006).

    Before the last 20 years only Johansen inseveral papers (Johansen 1982, Johansen1983a,b, 1986a,b) approached the topic of theuse of characters to rigorous analysis that showits validity in taxonomic decisions. Some otherworks have been strongest criticised for thechoice of the characters related with no correctthe hierarchical level or for the publication ofsome obscure cladistic results (Bhatti 1993).

    In several groups the comparative study of thecharacters in species that are close together isnecessary to establish the possible change of thestates of characters, but that is not enough forthe determination of the evolutionarydevelopment of each character (Wiley 1981,Retana-Salazar 2006).

    In the last years, important papers aboutmolecular data in Thysanoptera have appeared.One of the most important is the publication ofMorris and Mound (2005). Other researchersused morphological features as principalcharacters, but the use of morphologicalcharacters may be dangerous if the importanceof the character is not correctly evaluated asgood characters in phylogenetic resolution(Bjrklund 1997, Diniz Filho 2000, Retana-Salazar & Retana-Salazar 2008). In these casesthe subjective consideration of the possibleevolutionary origin of the characters is adangerous practice that shows the a priori ideasof the taxonomist more than the realevolutionary change.

    SYSTEMATIC PHYLOGENETICS IN THYSANOPTERA

    Mound and collaborators (1980) presented a phylogenetic paper concerning the familyclassification in Thysanoptera. This work hassome methodological problems that prevent therepetition of these results, especially the absenceof the type of analysis formally used, i.e. somedetails like the selected software, the

    polarization criterion of characters, the analysisof homoplasius and homologue characters, thestatistics of the trees obtained and the supportindex, among other things. For this reason theworks of Johansen (1982, 1986a,b) areconsidered the first formal phylogeneticanalysis. Excellent revisionary work was

    performed by this author on the taxonomy, biogeography and phylogeny of Elaphrothrips , Leptothrips and Humboldthripini. In the first paper about Humboldthripini this authorincludes all the tribes of Thripinae.

    Johansen is clear in the description of charactersand the detail of the methodology used for the

    phylogenetic reconstruction. This technique isthe only weakness of this analytical work,

    because it does not include the use of specificsoftware nor the analytical study of characters,

    but presents a good statistical approach to thedistribution of characters that permits therepetition of these results. However thedescription of characters is the most important,with extensive character states all well-defined.That is the real strength of this work andcharacter definition is the most important step ina good phylogenetic analysis.

    The work performed by Mound andcollaborators (1980) as a phylogenetic analysiswith absence of all the necessary technicalconsiderations, which are indispensable forothers that wish to try and reproduce thoseresults, is an example of some critical papersthat need extensive revision on the phylogeneticresults performed. Several times I have tried toutilize this data matrix with different softwarefor phylogenetic reconstruction but I havent

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    3/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    18

    been successful in obtaining the same results asthese authors (Figs 1-3). The only alternative inobtaining these results is using constraints forthe topologies depicted (Figs. 4, 5). This impliesthat these results are not in the set of optimalsolutions and is a particular restriction for a

    particular result. It is relevant to note that theresults obtained without the use of constraintsshow better statistical fit (Figs. 1-3). Theseresults indicate that the considerations of Bhatti(1993) seem to be correct when he commentsabout this work : The changes they prop osed inthe family classification of Terebrantia did notarise from cladistic analysis. They did notreconstruct the phylogeny. They merely plottedthe character states on two proposedcladograms(page 99). A similar problem isevidenced when analyzing the data matrix usedfor the study of the phylogenetic relationshipsamong groups of Aeolothripidae (Marullo &Mound 1995)(Figs. 6, 7).

    Recent works from Morris and collaborators(1999) have several methodological problemswidely discussed in phylogenetic papers. Theseauthors use very small numbers of charactersconsidering the number of taxa included. Thecharacters considered are not the best at ageneric and supraspecific level. Severalconclusions are not well founded in these resultsand they considered this phylogenetic analysisas definitive in order to propose patterns ofevolution.

    The works of Retana-Salazar (1998, 2000, 2001,2006, 2007a,b, 2010) are classical cladisticwork, using new alternatives in the phylogeneticreconstruction, particularly useful in highhomoplasius groups, where the common use of

    parsimony may conduce to unreal results(Brooks & Wiley 1986, Retana-Salazar &Retana-Salazar 2008). All propose newhypothesis and as such are subject to proof. Asrequired by formal science. Recently Cavallieriand Mound argued that As in all branches of

    science, conclusions in taxonomy can be nomore reliable than the data on which they are

    based (Cavallieri & Mound 2012, page 1).However, in many of Mounds articlesconcerning aspects of the phylogeny ofThysanoptera, there is a lack of reliability in themethod.

    These papers are an effort in the search ofobjective tools for the determination of thecharacters that show natural groups inThysanoptera (Retana-Salazar & Retana-Salazar2008). Based in his results this authordemonstrates the necessity to use phylogeneticanalysis to define the possible directionality ofevolutionary changes and the necessity tosegregate some genera into several ones butdefining the natural groups using phylogeneticanalysis (Retana-Salazar 2006, Retana-Salazar& Soto-Rodrguez 2007, Retana-Salazar &

    Nishida 2007).

    CLADISTICS AS A TOOL FOR TAXA DESCRIPTION

    The philosophy of cladistics suggests thenecessity to take some axiomatic purposes forthe correct reconstruction of phylogeny.

    Some particular groups as Frankliniella , Thrips , Hoplothrips , Elaphrothrips, Heterothrips andothers, have a great number of species and thehomoplasious level is high, commonobservation is not enough to define whichcharacters are homologues or homoplasious.The use of a data matrix of these characters toobtain a hypothetical tree permits the

    phylogenetic study of the characters behaviourand its phylogenetic value.

    Choosing a correct out-group may be difficultwhen the in-group is highly homoplasious. Thechoice of an incorrect out-group can lead toequivocal and not congruent relationships. Isuggest two practices for this delicate scenario:a) the use of several out-groups all of them with

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    4/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    19

    characters shared with the in-group (Retana-Salazar 2010), or 2) the use of a hypotheticalancestor created using characters observed inseveral other taxa (Retana-Salazar 2007a). Inthis aspect Mound and collaborators (1980) andBhatti (1988, 1993) suggest possible data for theinference of the hypothetical ancestor. The useof multiple out-groups can be very useful ingroups with few species where the diagnosticcharacters are well defined (Retana-Salazar2007a). By contrast, in taxonomic groups wherethere is no clarity on the limits of the taxa andthere is no homogeneity of the characters usedin grouping, the use of multiple out-groups maynot be of much help except to determine the

    paraphyly of the ingroup (Retana-Salazar 2010).

    BIOLOGICAL CHARACTERS IN THYSANOPTERA

    Some authors suggest (Mound & Marullo 1996)that the uses of these criteria are in several cases

    better than the morphological inference, becausethe plasticity of morphological features maylead us to mistaken associations. Lamentablythese characters are diagnostic in few cases.One relevant case where biology has been themost important feature is Aulacothrips recognized as family level by Bhatti (2006) it iswell supported by morphology, biology and

    biogeographic features. In this particularspecies, the only one known to be anectoparasite of insects and whos natural hostsare the individuals of Aetalion , and somemembracids is one extreme case and there is nodiscussion about this. Other groups are not soclear in those biological features (Alves-Silva &Del Klaro 2011).

    Several species of Frankliniella are closelyrelated to Asteracea species, and only one isstrictly associated with Hemerocallis , while feware strictly associated with mosses (Mojica &Johansen 1990, Mound & Marullo 1996,Retana-Salazar 1998), but it is difficult toestablish that these particular groups are

    different taxa only based on their biology.Retana-Salazar (2010) shows that these

    biological features are only significant for thecharacterization of species, but not for groups ofspecies.

    Biology may be highly homoplasious because biological features are very plastic, and in thesecases the uses of biological features in absenceof other morphological characters is not a goodindex for group segregation.

    MORPHOLOGICAL CHARACTERS AND THEPROBLEM WITH SPECIES RECOGNITION

    The morphological characters may be veryvariable, but we need critical studies of thegroup variation for making good taxonomicdecisions. Bhatti (2006) established thattaxonomy as we know it is reallymorphotaxonomy, because we need particularmorphological features for the description andcomparison between species.

    Some particular features as the brachipterouscondition of some species in Frankliniella andThrips may lead to the recognition of naturalgroups with a particular morphological feature(possibly apomorphic character) that defines thegroup. This consideration without a formalanalysis of other characters may bring us toincorrect phylogentical assumptions.

    In the genus Thrips it seems that this characteris associated with other derived characters andthe brachipterous species may form a realnatural group. In Frankliniella the situation isvery different and the brachipterous forms arenot supported by other derived characters andthe reduction of the wings seems to appearseveral times in the evolutionary history of thegroup (Retana-Salazar 2010).

    The number of antennal segments is analyzed by Mound and Palmer (1981) in six genus-

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    5/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    20

    groups in Thripina. This author suggested thatthe plesiotypic condition is 9 segments, somegroups in Aptinothripina and all the genera inPseudothripini seem to have a secondarydevelopment of this 9-segmented condition.Mound and Palmer (1981) suppose that thiscondition derives from the subdivision of thesixth segment. With this evidence Mound andPalmer (1981) suggest that the most commonapomorphic condition is seven antennalsegments. This consideration leads to the

    possibility of interpreting the presence of six,seven and eight as an apomorphic sequence.

    Recent studies in the genus-group Anaphothrips (Retana-Salazar 2007b, Retana-Salazar 2010)demonstrate that these conditions may bereversal on several occasions. They are useful inthe diagnosis of genera, the studies in the

    Anaphothrips group evidenced that thistransformation has good phylogenetic inertiawith the hierarchical level of genus. A similarcondition is presented by the phylogeneticanalysis of the Hoodothrips genus-group(Retana-Salazar 2007a). In this particular casethe fusion of segments may occur not only inthe terminal segments. This feature complicatesthings when considering apomorphic and

    plesiomorphic conditions. Those are goodexamples of how the same characters, in generathat seem phylogenetically close, or at leastincluded in the same family, have a verydifferent evolutionary landscape.

    Mound and Palmer (1981) proposed the uses of8 morphological characters for the segregationof genus-groups in Thripina. These authors

    proposed a particular polarity from thesecharacters based on merely observational dataof slides. This methodology suggested that theanalysis may be erroneous; Retana-Salazar(2008) uses a data matrix with these 8characters in the phylogenetic reconstruction ofthe relationships of the genus-groups proposed

    by Mound and Palmer (1981). In this paper

    Retana-Salazar (2008) considered thosecharacters under a formal analysis and use forthe determination of polarity recently describedfossil form with a good number of specimensthat permit a good observation of the major partof characters (Pealver 1998). Results of thisanalysis show that the values of these charactersin genus-groups definition are very poor and themajority is not well associated with the

    phylogenetic structure of these groups. Thiskind of work reveals that considerations about

    polarity of characters in absence of formalanalysis of phylogeny and based only inobservation of the alpha taxonomists, may leadto grave mistakes that may be reflected inerroneous taxonomic decisions.

    For this reason we need to 1) establish a phylogenetic hypothesis and 2) decide based onthese results which characters have a good

    phylogenetic fit in each case and then these will be the features used in the definition of naturalgroups. This fit needs to be measured usingsome important tools, like the measure of the

    phylogenetic inertia of characters (Bjrklund1997, Diniz-Filho 2000, Retana-Salazar &Retana-Salazar 2008) and the evaluation of theless entropic topology for a same data matrix(Brooks & Wiley 1986, Retana-Salazar 2007a,Retana-Salazar & Retana-Salazar 2008).

    In Tubulifera the problem is the same. Moundand Marullo (1996) proposed that in the

    Liothrips lineage, the presence of one sensoriumin antennal segment III and three sensoria inantennal segment IV are characteristics of thisgeneric group. The phylogenetic analysis

    performed by Morris and collaborators (1999)using morphological features shows that thiscondition in antennal segments III-IV came upseveral times in the evolutionary history of thegroup. These characters have a low

    phylogenetic inertia and for this reason are notgood in defining the lineage.

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    6/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    21

    However this kind of characters is good for thesegregation of generic taxa, particularly thenumber of antennal sensoria in antennalsegment IV in the Karnyothrips lineage(Retana-Salazar & Soto-Rodrguez 2007).Recently Minaei and collaborators (2007)redefined the genus Noeheegeria , and in this

    paper, without any particular phylogeneticarrangement the authors considered that thisfeature (antennal sensoria), is indeed valid forthe segregation of genera. These authorsanalyzed the presence of this feature in a groupof species that shows this particular character,

    but its presence is not enough for separating ifwe do not know whether these characters are

    plesiotypic or apotypic, because apormorphouscharacters could support all segregation. This

    biological problem to define which charactersare apomorphous and which of them are

    plesiotypic, as well as the determination ofhomoplasius characters, is only resolved by

    phylogenetic analysis.

    These problems with morphological charactersare biological hurdles usually appointed inliterature. Moreover, there are other kinds of

    problems with morphology in Thysanoptera.The preparation procedure can affectmorphology. For example, Hoyers mediumcreates undulating setae and produce distortionof the colour and the original shape of thespecimen. This is due to it not being viscous atall and the cover slide then stretches the sample,destroying the original shape of the specimen.The macerating substance affects coloration andthe clarity with which some characters areobserved. Excess pressure on the cover slidestretches the sample producing distortion of themorphological features, for example widerabdomen and head bases, together with longermouth parts and displaced and longer setae(Retana-Salazar & Mound 1994).

    Retana-Salazar & Mound (1994) proposed thatvariation within a species can be greater than

    variation among several species, but thisaffirmation needs further studies because highlyvariables species such as Fr. occidentalis and

    Fr. intonsa may be a species complex. Mound(2005) suggests that phenotypic plasticity is agreat source of morphological variation withinspecies, but this asseveration needs severalcontrolled experiments to prove it and not justthe personal guess of a taxonomist.

    MORPHOLOGICAL TRAITS OF THE IMMATURESTAGE

    The characters of the immature stages could beuseful determining the species as in the case of

    Selenothrips rubrocinctus where the speciesname refers to the main feature of the larva II,the presence of a red band at the middle of

    body. In other cases, such as Taeniothripsinconsequens where has been reported the

    presence of a well-developed toothed comb inthe terminal segments of the abdomen of thelarva II, which is characteristic of the species(Lacasa and Llorens 1996)(Fig. 8). The

    presence of distinctive characters in somespecies has caused to put aside the integral

    description of the larva II, thereby losing a largeamount of information that can be of value instudies of phylogeny.

    The characters of the immature can be useful in phylogeny and taxonomy, making it necessaryto develop studies of the ontogeny of thecharacters, as has been done in groups likeCulicidae (Harbach 1991). In Thysanopterahave been studied little morphologicalcharacters of the different stages of

    development. There is no clear characterizationof the genera at the immature level but if thereis numerous descriptions of the larvae of somespecies of thrips. A first step in systematicresearch into the characters of larval forms inThysanoptera is to determine the generalfeatures that define the different genera andtaxonomical value of these characters.

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    7/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    22

    An interesting example is the study of theimmature of Frankliniella insularis where theadult female is characterized by the absence ofmedial comb on the posterior margin of sternumVIII (Fig. 9a), among other features, but theobservation of larvae II shows that this comb isnot present in abdominal segment VIII but asimilar but small and complete similar structureis in abdominal tergum IX in this stage (Fig.9b). This can be useful in the polarization of thischaracter.

    The structure of the segments of the antenna can be a good Terebrantia taxonomic character. InThripidae segments III-IV of immatures are

    very developed and segments V-VII are forminga stylus (Fig. 10a, b). In Aeolothripidae antennalsegments are evident since very early as welldeveloped units of similar structure to that ofadults but with fewer segments (Fig. 11a). Inaddition, in Aeolothrips intermedius hasascertained the presence of a double comb in thecaudal segments of the abdomen (Fig. 11b),while in adults there is no evidence of suchstructures.

    Preliminary data has given rise to start at theUniversity of Costa Rica a project to describesystematically the characters of the larvae of the

    principal species of thrips. These data will beginstudies on the importance of ontogeny in

    phylogenetic characters polarization.

    TERATOLOGICAL CHARACTERS

    These characters are abnormalities recorded bytaxonomists after the revision of a long series of

    individuals. Such variation is highly diverse, inmost animals it may be identified because theindividuals involved either deviate markedlyfrom type as to be recognized as freaks or

    because the abnormalities involved areasymmetrical (Mayr 1969). Teratologicalcharacters may not be used in taxonomicdecisions, however recently Goldarazena and

    collaborators (2008), used a particular characterin a few individuals to synonymize a recentlydescribed species. This type of mistake may bedangerous because some journals publish theseequivocal papers. Asiain and Mrquez (2009)show that several specimens in some particularColeoptera groups, are highly asymmetricallymodified and for this reason asymmetries arenot good characters for taxonomic work.Asymmetry is a good indicator that thischaracter is not possible to be valuated to

    propose taxonomic changes (Mayr 1969).Cappe de Baillon (1927) and Balazuc (1948)studied and classified the teratological variation.Recent studies in highly polluted areas ofEurope has shown that some sensitive specieslike Frankliniella intonsa may be drasticallyaffected promoting an increase in the rate ofthese abnormalities and the majority of thesemonstrosities are associated with antennaldeformations (Vasiliu-Oromulu et al. 2008).

    However, the abnormalities reported as a resultof environmental issues, reflected inembryological deformations in species such as

    Frankliniella intonsa , are also found inspecimens collected in undisturbed systems intropical regions. This is the case of some speciesof Aeolothrips and Liothrips collected in CostaRica and in some specimens of Erotidothripsmirabilis collected in Asian national parks. It isnecessary, as has been done in other insectsgroups, to begin producing literature that reportsand illustrates these morphological changeswhich may lead to taxonomic errors inThysanoptera.

    MOLECULAR DATA IN THYSANOPTERA

    In several cases this kind of data are consideredthe panacea for all the taxonomic and

    phylogenetic problems. But the real landscape isvery different. In some aspects molecular dataare more objective than traditional characters,

    but in other aspects these molecular characters

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    8/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    23

    are inherently subjective (Lee 2004). Theconsiderations of some specialists that statemolecular data as being the solution to our

    problems only denote little knowledge aboutthis topic.

    Frequently the molecular data are not consistentwith the resolution. Same sets of data showvery different solutions and the question is whatis the truth? In reality, all the answers are true

    because the inconsistency of the resolution ofdifferent sets of data is an effect of differentmethodologies for the study of the data matrix(Kjer et al. 2006, Klass 2007). In other casesthe molecular data concludes with resolutionswithout consistency with the biology of thegroup (Mound & Morris 2007)

    Mound and Morris (2007) used molecular datato reconstruct the phylogeny of theThysanoptera. The only clear result that theseauthors obtained was the monophyly ofThysanoptera. The only arrangements withsupport for natural groups definition wasTubulifera and Terebrantia, the same twogroups defined by Haliday in 1836. The familygroups are not well supported and the classicalfamilies are not consistent natural groups. Theseresults are totally inconsistent with the

    biological and morphological features. Severaltimes molecular studies have the sameimportance than a study in the Amazonia toconclude that this is a forest.

    On the other hand, the molecular data needsgood support in morphological featuresassociated with the phylogenetic results,otherwise these results are inconsistent. Crespiand collaborators (2004) with molecular datamatrix conclude that in some groups as theclade formed by( Advenathrips +Vicinothrips )( Koptothrips (Triad othrips (Crespithrips (Glaridothrips + Xaniothrips))))) in the basal node of Koptothrips theenlarged forelegs character appears, which may

    diagnose this group, but this character is notconstant in the species of this group. Studiesconducted in other arthropod groups haveconcluded that molecular data may indicate the

    presence of genetic units with a high rate ofgenetic divergence which could indicate the

    presence of undescribed species. However, it isunclear what level of divergence is needed forallowing the delimitation of two lineages(Camacho et al. 2011).

    CONCLUSIONS

    Thysanoptera is one of the highesthomoplasious groups such as Hymenoptera andHemiptera (Nieto 1999, Gauld & Mound 1982,Mound & Palmer 1983). Mound supposes thatthe operational difficulty with systematic studieson Phlaeothripinae is that many taxa at genericlevel and above are defined by loss ofapomorphous characters.

    Gauld & Mound (1982) conclude that genera ingroups of insects with high levels of homoplasyneed to be defined polythetically, which isthrough the presence or absence of one or moreunique characters. About this point Mayr(1969) proposed three statements that need to besatisfied for a polythetic taxon: a) each species

    possesses a large number of the total number ofcharacters of the taxon, b) each character is

    possessed by a large number of the species andc) no characters are possessed by every speciesof the aggregate but is missing in the species ofall the other taxa.

    The polythetic taxa created by Mound as Holopothrips does not follow the anteriorstatements and these genera defined in absenceof real characters are a big black box where the

    biology and morphology needs to be reviewedfor a proper taxonomy of these species (Retana-Salazar & Nishida 2007).

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    9/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    24

    In the last years some authors considered thatgenitalia is not highly informative in thripstaxonomy, and only some structures highlyvariables as pseudovirga are more or less useful(Mound & Minaei 2007). This consideration isopposite to other data about morphologicalvariation derived from analysis of long series ofspecimens and the analysed together with thegeographic variation (Retana-Salazar 2006,2009). These opposite results are indicators ofthe complexity of thrips characters. Despite theconsiderations regarding the observations ofsome taxonomists, there is a big amount ofliterature that demonstrates the utility of themale genitalia in taxonomic works (Crdoba-Aguilar 2000).

    Molecular data in Thysanoptera is not good fortaxonomic decisions and makes for poorcharacters in order to define biological groups(Mound & Morris 2007). Thysanoptera likeother groups, has several problems if moleculardata are considered as the best solution totaxonomic and phylogenetic problems.Molecular data is not the solution to taxonomicissues and the objectivity in some aspects of thiskind of data is accompanied by the subjectivityin other considerations (Lee 2004). The abilityof molecular data to determine species bygenetic similarity, remains valid if the species ofthe group are well known, at the same time, the

    possibility of using molecular data to identifynew species appears to be an unreliabletechnique as well, if the data-bases are not largeenough or if taxonomic confirmations of theirwork are lacking (Meyer & Paulay 2005).

    Teratological studies are necessary inThysanoptera because some erroneoustaxonomic decisions have been proposed usingthis kind of characters. In some other groups ofinsects these particular characters have beenstudied for the taxonomic significance of these(Asiain & Mrquez 2009). Goldarazena andcollaborators (2008) synonymyze the species

    Nicolemma garitai with Aurantothripsorchidaceus based on teratological charactersand these decisions are not well founded andlead to grave mistakes in taxonomic work.These kinds of publications in sections used in

    journals for discussion of particular areas ofconflict with few arguments destroy goodtaxonomic proposals for the sole comfort oftaxonomists.

    We need a natural classification ofThysanoptera, but this group is highlyhomoplasius, the alternative is to carry out

    phylogenetic studies that help us in thedefinitions of higher taxa and the study of the

    phylogenetic inertia associated to characters thatshow us which features are good for thedefinition of natural groups that may beconsidered as taxa.

    Several times specialists consider that thecharacters that are good in a particular groupmay be extrapolated to other taxa. In highlyhomoplasius groups this a priori consideration may conduce to consider as validartificial groups with homoplasius charactersused for their diagnosis.

    Mound and Palmer (1981) published aninteresting paper where they analyzed eightmorphological characters in 50 genera ofThripini, they conclude that these features havea high taxonomic value and propose the

    polarization of these characters based only onthe mere observation of specimens and inabsence of formal analysis. Based on thesecharacters these authors propose some genusgroups Frankliniella genus group, Thrips genusgroup, Taeniothrips genus group, Mycterothrips genus group and other. Recent studies based in

    polarity of characters using fossil data and anew methodology to study the fossil insertion in

    phylogenetic trees show that these charactersare not good for defining genus groups and arenot informative. Only two characters are useful

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    10/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    25

    for phylogenetic structure and both are plesiotypic and consequently not good fordetermination of natural groups (Retana-Salazar2008). That is an example of the problemscarried on with classification systems based ontaxonomic observation of characters in absenceof formal phylogenetic analysis.

    Pealver and Nel (2010) have recently published an excellent paper about thetranscendent importance of the fossil evidencein determination of the contribution ofmorphological characters and based in theevidence and a new and better conducted

    phylogenetic analysis conclude that the familyStenurothripidae needs to be considered a goodtaxon and an acceptable name better thanAdiheterothripidae. More evidence is needed tosupport this new hypothesis but at least this

    proposal is solidly founded in data and formalanalysis. Previous studies have spurious

    phylogenetic analysis where the repetition of theanalysis shows a different result to those thathave been published (Mound & Marullo 1998).

    The only real tool to define hypothetical naturalgroups, based on the presence of apomorphouscharacters, is the phylogenetic reconstructionand the analysis of the phylogenetic inertia ofthe characters, that permits the consideration ofwhich characters are better to be used intaxonomic work. There arent any goodindicators for phylogenetic inertia of charactersin qualitative data sets, but some authors have

    proposed the use of the retention index for thisevaluation (Bjrklund 1997, Diniz-Filho 2000).

    Several times, with high complex groups likeThysanoptera, the parsimony analysis is not agood tool because the results of thereconstruction are a false clade, for this reasonBrooks & Wiley (1986) suggest the uses of theentropy concept in phylogeneticreconstruction as a way of evaluation of the

    trees obtained. Recently Retana-Salazar &Retana-Salazar (2008) published a book wherethe uses of entropy in the study of biologicalsystems are proposed, particularly theevolutionary systems and the determination ofnew taxa.

    ACKNOWLEDGEMENTS

    To Ral Ramrez-Morales and Stephen Smithfor the English revision of this paper andAlexander Rodrguez-Arrieta for his commentsto this paper, and also The project Descripciny ultraestructura de los thrips (Thysanoptera:Insecta) de Mesoamrica 810 -A6-239 and the

    project "Estudio morfolgico y gentico de losestados inmaduros de thrips (Thysanoptera:Insecta) de relevancia econmica enHispanoamrica", N810-Bl-224. Thecollections of the British Museum of NaturalHistory, London, England, SenckenbergMuseum, Frankfurt, Germany, MuseoGeominero de Madrid, Espaa, Collection deThysanoptera du Muse d'Histoire Naturelle deParis, National Collection of South Africa,Pretorya, South Africa, IBUNAM, Coyoacan,Mexico DF, Universidad Autnoma de laHabana, Cuba, Centro Nacional de SanidadVegetal de Cuba, Universidad AutnomaAgraria Antonio Narro, Coahuila, Mexico,Universidad Autnoma de Nayarit, Nayarit,Mexico, INBio, Santo Domingo de Heredia,Costa Rica. Alcides Snchez-Monge foroutstanding production of SEM images.

    R EFERENCES

    Alves-Silva, E., Del Klaro, K. 2011. Ectoparasitismand phoresy in Thysanoptera: the case of

    Aulacothrips dictyotus (Heterothripidae) in the Neotropical savanna. Journal of Natural History 1-13.

    Balazuc J. 1948. La teratologie des Coleopteres etexperiences de transplantation sur Tenebrio

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    11/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    26

    molitor L. Memoires de Meseum NationaldHistoire Naturelle 25:1-293.

    Bhatti JS. 1993. Phylogenetic relationships among

    Thysanoptera (Insecta) with particular reference tothe families of the Order Tubulifera. Zoology(Journal of Pure and Applied Zoology) 4:93-130.

    Bhatti JS. 2006. The classification of Terebrantia(Insecta) into families. Oriental Insects 40:339-375.

    Bjrklund M. 1997. Are comparative methodsalways necessary? Oikos 80:607-612.

    Brooks DR, Wiley EO. 1986. Evolution as Entropy:Toward a Unified Theory of Biology, ChicagoUniv. Press.

    Camacho, A.I., Dorda, B.A., Rey, I. 2011.Identifying cryptic speciation across groundwater

    populations: first sequences of Bathynellidae(Crustacea, Syncarida). Graellsia 67(1):7-12.

    Cappe de Baillon P. 1927. Recherches sur alteratologie des insectes. EncyclopediedEntomologie 8:5-291.

    Cavalleri, A., Mound, L.A. 2012. Toward theidentification of Frankliniella species in Brazil(Thysanoptera, Thripidae). Zootaxa 3270:1-30.

    Crdoba-Aguilar A. 2000. Evolucin y diversidadde la morfologa de los genitales masculinos eninsectos. Folia. Entomol. Mex. 110:95-111.

    Diniz-Filho JAF. 2000. Mtodos FilogenticosComparativos. Holos Editora. Brasil.

    Gauld ID, Mound LA. 1982. Homoplasy and thedelination of holophyletic genera in some insectgroups. Systematic Entomology 7:73-86.

    Goldarazena A, Mound LA, zur Strassen R.2008. Nomenclatural problems amongThysanoptera (Insecta) of Costa Rica. Revista deBiologa Tropical (Forum) 56(2):961-968.

    Harback, RE. 1991. Ontogeny of the larval stage ofSabethes chloropterus, with special reference tosetal development and phylogenetic implications for the family Culicidae (Diptera). Mosquito

    Systematics 23(1):10-24

    Johansen RM. 1982. Algunos aspectos de la biologa, ecologa, conducta y distribucin degeogrfica del gnero Torvothrips (Insecta:Thysanoptera). Anales del Instituto de Biologa.Universidad Nacional de Mxico 52:205-222.

    Johansen RM. 1983a. El gnero Elaphrothrips Buffa, 1909 (Thysanoptera: Phlaeothripidae) en elcontinente Americano; su sistemtica, evolucin,

    biogeografa y biologa. Monografa del Instituto

    de Biologa, Univeridad Nacional de Mxico 1:1-267.

    Johansen RM. 1983b. Nuevos thrips (Insecta:Thysanoptera; Terebrantia, Thripidae, Thripinae)de La Sierra Madre Oriental y del Eje VolcnicoTransversal de Mxico. Anales del Isntituto deBiologa, Universidad Nacional de Mxico 53:91-132.

    Johansen RM. 1986a. Revisin de la tribuHumboldthripini Johansen, 1983 (Insecta;Thysanoptera; Thripinae). Anales del Isntituto deBiologa, Universidad Nacional de Mxico 56:697-744.

    Johansen RM. 1986b. Nuevos conceptostaxonmicos y filogenticos del gnero

    Elaphrothrips Buffa, 1909 (Thysanoptera:Phlaeothripidae) del continente americano ydescripcin de dos especies nuevas. Anales delInstituto de Biologa, Universidad Nacional deMxico 56:745-868.

    Lacasa Plasencia A. y Llorens Climent, J.M.1996. Thrips y su control biolgico. DivulgacinTcnica 18, Consejera del Medio Ambiente,Agricultura y Agua, Murcia, Espaa 312pp.

    Lee M. 2004. The molecularisation of Taxonomy(Review). Invertebrate Systematics 18:1-6.

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    12/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    27

    Marullo R, Mound LA. 1995. Su unaclassificazione sopra-generica della famigliaAeolothripidae (Thysanoptera). Atti del XVIICongresso Nazionale Italiano di Entomologia,

    Udine: 87 90.

    Mayr E. 1969. Principles of Systematic Zoology.McGraw-Hill Inc. New York. p 26.

    Meyer CP, Paulay G. 2005. DNA barcoding: Errorrates based on comprehensive sampling. PLoS Biol3(12): e422.

    Minaei K, Azemayeshfard P, Mound LA. 2007.The southern Paleartic genus

    Noeheegeria (Thysanoptera:Phlaeothripidae):redefinition and key to species. Tijdschrift loorEntomologie 150:5-64.

    Mojica-Guzmn A, Johansen RM. 1990. Estduiosucesional de tisanpteros (Insecta), habitants delquenes y musgos, en cinco localidades de laSierra Madre Oriental, Estado de Hidalgo, Mxico.Anales del Isntituto de Biologa, Universidad

    Nacional de Mxico 61:197-256.

    Morris DC, Mound LA, Scwarz MP, Crespi BJ.1999. Morphological phylogentics of Australinagall-inducing thrips and their allies: the evolutionof host-plant affiliations, domicile use and social

    behaviour. Systematic Entomology 24:289-299.

    Mound LA. 2005. Fighting, fligth and fecundity: behavioural determinants of Thysanopterastructural diversity.pp 81-105 in Ananthakrishnan,TN & Whitman D. (eds). Insects and phenotypic

    plasticity. Science Publishers Inc. Enfield, NH,USA.

    Mound LA, Palmer JM. 1983. The generic andtribal classification of spore-feeding Thysanoptera(Phlaeothripidae: Idolothripinae) 46(1):1-174.

    Mound LA, Heming BS, Palmer JM. 1980.Phylogenetic relationships between the families ofrecent Thysanoptera (Insecta). Zoologycal Journalof the Linnean Society 69:111-141.

    Mound LA, Palmer JM. 1981. Phylogeneticrelationships between some genera of Thripidae(Thysanoptera). Entomologica Scandinava Suppl.15:153-170.

    Mound LA, Marullo R. 1998. Two new basal-cladeThysanoptera from California with Old Worldaffinities. Journal of the New Yoek EntomologicalSociety 106:81-94.

    Mound LA, Minaei K. 2007 . Australian thrips ofthe Haplothrips lineage (Insecta: Thysanoptera).Journal of Natural History 41(45-48):2919-2978.

    Mound LA, Morris DC. 2007. The Insect OrderThysanoptera: Classification versus Systematics.Zootaxa 1668:395-411.

    Pealver-Moll E. 1998. Estudio taxonmico y paleoecolgico de los insectos del Mioceno deRubielos de Mora (Teruel). Instituto de EstudiosTurolenses, Excma. Diputacin Provincial deTeruel, Espaa. 177pp.

    Pealver-Moll E. & Nel P. 2010. Hispanothrips from early Cretaceous Spanish amber, a new genusof the resurrected family Stenurothripidae (Insecta:Thysanoptera). Ann. Soc. Entomol de France (n.s)46(1-2):138-147.

    Retana-Salazar AP.1998a. Reestablecimiento delos gneros Frankliniella , Exophthalmothrips y

    Bolbothrips (Thysanoptera: Thripidae) Revista deBiologa Tropical 46(2):385-396.

    Retana-Salazar AP. 1998b. Una visin filogenticade Frankliniella (Thysanoptera: Thripidae).Revista de Biologa Tropical 46(2):397-406.

    Retana-Salazar AP. 2000. Revisin y Filogenia delgrupo genrico Pseudothrips (Thysanoptera:Thripidae). BRENESIA 54:51-62.

    Retana-Salazar AP. 2006. Filogenia y revisiontaxonmica de la superfamilia Pediculoidea n.stat.(Anoplura: Insecta). BRENESIA 66:15-24.

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    13/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    28

    Retana-Salazar AP. 2007a. El grupo genrico Hoodothrips (Terebrantia: Heliothripidae). ActaZoolgica Lilloana 51(1):15-38.

    Retana-Salazar AP. 2007b. Los tisanpteros delgrupo genrico Anaphothrips (Thysanoptera:Thripidae), conenfasis en Amrica Central. Revistade Biologa Tropical 55(1):321-333.

    Retana-Salazar AP. 2010. El grupo genrico Frankliniella : el significado filogentico de sus principales caracteres morfolgicos (Thysanoptera:Thripidae; Thripini). Mtodos en Ecologa ySistemtica 5(3):1-22.

    Retana-Salazar AP, Mound LA. 1994. Thrips ofthe Frankliniella minuta group (Insecta:Thysanoptera) in Costa Ricen Asteracea flowers.Revista de Biologa Tropical 42:639-648.

    Retana-Salazar AP, Ramrez-Morales R. 2006.Establecimiento de un nuevo gnero de piojos(Phthiraptera: Pediculidae) asociado al hombre(Primates: Hominidae). BRENESIA 65:61-70.

    Retana-Salazar AP, Nishida K. 2007. First gall-inducing thrips on Elaphoglossum ferns: A newgenus and species of thrips Jersonithrips

    galligenus from Costa Rica (Insecta, Thysanoptera,Phlaeothripidae). Senckenbergiana biologica87(2):143-148.

    Retana-Salazar AP, Soto-Rodrguez GA. 2007.Revisin taxonmica del grupo Haplothrips -

    Karnyothrips (Thysanoptera: Phlaeothripidae).Revista de Biologa Tropical 55(2):627-635.

    Retana-Salazar AP. 2008. Un enfoque cladista parala insercin de fsiles en la filogenia. RevistaGeolgica de Amrica Central. 39:93-106.

    Retana-Salazar AP, Retana-Salazar SA. 2008.Entropa Biolgica. Especies y Filogenia. Editorialdel Instituto Centroamericano para la InvestigacinBiolgica y la Conservacin (ECIBRC). San Jos,Costa Rica 167p.

    Vasiliu- Oromulu L, Jenser G, Brbuceanu D.2008. Frankliniella intonsa (Trybom, 1895) a verysensitive bioindicator for air pollution. ActaPhytopatologica et Entomologica Hungarica,

    43(1): 401-408.

    Wiley EO. 1981. Phylogenetics. The Theory andPractice of Phylogenetic Systematics. New York:Wiley-Interscience.

    Received: 10 de Diciembre 2011Reviewed: 03 de Febrero 2012Accepted: 23 de Agosto 2012.

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    14/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    29

    Figure 1. Topology obtained by applying a heuristic analysis of the data matrix presented by Mound et al.1980. Statistics L=144, CI=0.94, R=0,65, analysis performed with PAUP 3.1.1.

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    15/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    30

    Figure 2. Topology obtained by applying a heuristic analysis of the data matrix presented by Mound et al.1980. Statistics L=144, CI=0.94, R=0,65, analysis performed with PAUP 3.1.1.

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    16/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    31

    Figure 3. Best topology obtained by applying an exhaustive analysis of the data matrix presented by Moundet al. 1980. Statistics L=144, CI=0.94, R=0,65, analysis performed with PAUP 3.1.1.

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    17/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    32

    Figure 4. First topology proposed by Mound et al. 1980

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    18/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    33

    Figure 5. Second topology proposed by Mound et al . 1980

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    19/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    34

    Figure 6. Phylogenetic tree of groups of the Aeolothripidae published by Marullo and Mound (1995).

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    20/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    35

    Figure 7. Topology obtained by exhaustive analysis of the data matrix proposed by Marullo and Mound(1995). PAUP 3.1.1 was used.

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    21/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    36

    Figure 8. Taeniothrips inconsequens immature stage II abdominal segment IX, dorsal view. SEMmicrophotography, 300X, Hitachi 570S (scale: 5,5cm100m)

    A B

    Figure 9 . Frankliniella insularis . A) Adult female comb on VIII tergum, abdominal segments. SEMmicrophotography, 600X, Hitachi 570S (scale: 5,7cm50m). B) Immature stage II comb on IX tergum,abdominal segments. SEM microph otography, 600X, Hitachi 570S (scale: 5,7cm50m)

  • 8/13/2019 Characters in Thysanoptera Mes 7(2) 16-37

    22/22

    Mtodos en Ecologa y Sistemtica Vol. 7(2):ISSN impreso 1659-2182ISSN digital 1659-3049 Setiembre-Octubre 2012

    [email protected]

    37

    A B

    Figure 10. Thripidae antennal segments of immature stage II. A) Head and antennal segments I-II(Taeniothrips inconsequens , SEM microphotography, 500X, Hitachi 570S, scale: 5,7cm60m). B)Antennal segments III-VII ( Taeniothrips inconsequens , SEM microphotography, 700X, Hitachi 570S, scale:5,7cm43m).

    A B

    Figure 11. Aeolothrips intermedius . A) Antenna, immature stage II (SEM microphotography, 400X, Hitachi570S, scale: 5,7cm7 5m). B) Abdominal terga VIII-X, immature stage II (SEM microphotography, 400X,Hitachi 570S, scale: 5,5cm75m).