Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/221753640

FactorsInfluencingtheGeographicalDistributionofDendroctonusrhizophagus(Coleoptera:Curculionidae:Scolytinae...

ArticleinEnvironmentalEntomology·June2011

DOI:10.1603/EN10059·Source:PubMed

CITATIONS

17

READS

103

4authors,including:

Someoftheauthorsofthispublicationarealsoworkingontheserelatedprojects:

InferenceoftheevolutionofstrainsHelicobacterpyloriinchildrenwith

recurrentinfection.Viewproject

GerardoZúñiga

InstitutoPolitécnicoNacional

82PUBLICATIONS873CITATIONS

SEEPROFILE

AllcontentfollowingthispagewasuploadedbyGerardoZúñigaon24September2014.

Theuserhasrequestedenhancementofthedownloadedfile.

POPULATION ECOLOGY

Factors Influencing the Geographical Distribution of Dendroctonusrhizophagus (Coleoptera: Curculionidae: Scolytinae) in the Sierra

Madre Occidental, Mexico

MA. GUADALUPE MENDOZA,1 YOLANDA SALINAS-MORENO,1 ANTONIO OLIVO-MARTINEZ,2

AND GERARDO ZUNIGA1,3

Environ. Entomol. 40(3): 549Ð559 (2011); DOI: 10.1603/EN10059

ABSTRACT The bark beetle, Dendroctonus rhizophagus Thomas & Bright, is endemic to the SierraMadre Occidental (SMOC) in Mexico. This bark beetle is a major pest of the seedlings and youngsaplings of several pine species that are of prime importance to the nationÕs forest industry. Despitethe signiÞcance of this bark beetle as a pest, its biology, ecology, and distribution are poorly known.Three predictive modeling approaches were used as a Þrst approximation to identify bioclimaticvariables related to the presence of D. rhizophagus in the SMOC and to obtain maps of its potentialdistribution within the SMOC, which is a morphotectonic province. Our results suggest that the barkbeetle could have an almost continuous distribution throughout the major mountain ranges of theSMOC. This beetle has a relatively narrow ecological niche with respect to some temperature andprecipitation variables and inhabits areas with climatic conditions that are unique from those usuallyprevalent in the SMOC. However, the bark beetle has a broad ecological niche with respect to thenumber of hosts that it attacks. At the macro-scale level, theD. rhizophagus distribution occurs withinthe wider distribution of its main hosts. The limit of the geographical distribution of this bark beetlecoincides with the maximum temperature isotherms. Our results imply a preference for temperatehabitats, which leads to the hypothesis that even minor changes in climate may have signiÞcant effectson its distribution and abundance.

RESUMEN l descortezador, Dendroctonus rhizophagus Thomas & Bright, es una especie endemica dela Sierra Madre Occidental (SMOC) en Mexico, que coloniza plantulas y arboles jovenes (�3 m) de variasespecies de pinos de importancia economica para la industria forestal de Mexico. A pesar de la importanciade este escarabajo como plaga, su biologõa, ecologõa y distribucion son poco conocidas. En este estudiorealizaron tres aproximaciones de modelaje predictivo para identiÞcar variables bioclimaticas relacionadascon la presencia de D. rhizophagus en la SMOC y para obtener mapas de su distribucion potencial. Losresultados sugierenqueeldescortezadorpuededistribuirsedemaneracasi continuaa lo largode laSMOC,que su nicho ecologico es relativamente estrecho con respecto a algunas variables de temperatura yprecipitacion y que habita en zonas con condiciones climaticas particulares que no son las que prevalecenen la SMOC. Sin embargo, esta especie tiene un amplio nicho ecologico con respecto al numero dehuespedes que parasita. A nivel de macro-escala la distribucion de D. rhizophagus esta contenida dentrode ladistribucionmas ampliade sushuespedesprincipalesyel lõmitede sudistribuciongeograÞcacoincidecon las isotermas de temperatura maxima. Por ultimo, los resultados indican una preferencia por habitatstemplados, lo cual permite hipotetizar que aun cambios menores en el clima pueden tener un efectosigniÞcativo sobre su abundancia y distribucion geograÞca.

KEYWORDS bark beetles,Dendroctonus rhizophagus, geographical distribution, predictive models

Both environment and population ecology processeslimit the abundance and geographic distribution ofspecies (Brown 1984, Thomas et al. 2001, Lomolino et

al. 2006). Traditionally, distribution studies have usedonly basic information (location and elevation) asso-ciated with specimens from museums and scientiÞccollections (MacDonald 2003). Integrating this ba-sic information with additional data on climate, bi-ology, and topography is essential to fully under-stand the distributional ranges of species, thepotential for distributions to change, and the vari-ables that limit or favor range expansion or con-traction (Gaston 2003).

1 Laboratorio de Variacion Biologica y Evolucion. Escuela Nacionalde Ciencias Biologicas-IPN. Departamento de Zoologõa. Carpio y Plande Ayala s/n, Col. Santo Tomas, C. P. 11340 Mexico City, Mexico.

2 Sanidad Forestal, Comision Nacional Forestal, Seccion VI delNoroeste, Chihuahua Mexico. Avenida Universidad piso 1. No. 3705.C. P. 31170, Colonia Magisterial.

3 Corresponding author, e-mail: capotezu@hotmail.com.

0046-225X/11/0549Ð0559$04.00/0 � 2011 Entomological Society of America

The use of geographic information systems (GIS)and the development of a broad spectrum of spatialmodeling approaches, including bioclimatic analysis(BIOCLIM) (Nix 1986), Ecological Niche FactorAnalysis (ENFA) (Hirzel et al. 2002), and MaximumEntropy (MaxEnt) (Phillips et al. 2004), have madepossible the integration of diverse data on biology,climate, and topography. This has led to the develop-ment of more reliable distribution maps and the mod-eling of species responses to different past and futureclimatic conditions. BIOCLIM, in particular, has beenused to predict the climate domain and potential spe-cies response to climate change at the meso-scale level(Mackey and Lindenmayer 2001, Ganeshaiah et al.2003, Jimenez-Valverde et al. 2007). ENFA and Max-Ent have been used to characterize ecological nichesand to determine suitable areas for species conserva-tion and natural resource management (Braunischand Suchant 2007, Sattler et al. 2007, Titeux et al. 2007,Kumar and Stohlgren 2009). Frequent use of thesetools has shown their usefulness and reliability in an-alyzing the potential distribution of species that havea relatively wide distribution range and whose biologyand ecology are poorly known (Pearson and Dawson2003, Beaumont et al. 2005).

Bark and ambrosia beetles (Coleoptera: Curculion-idae: Scolytinae) are important natural components offorest ecosystems. The infestation of weakened ma-ture trees promotes the presence of different succes-sional stages and demographic structures within theforest. Nevertheless, these insects are a worldwideproblem for the forest industry and in public parks andnatural reserves because of the extensive tree mortal-ity they cause under outbreak conditions and as vec-tors of diseases that reduce tree vigor (Wood 1982).Moreover, some bark beetles are important invadersthat represent a signiÞcant threat to the health offorests (Brockerhoff et al. 2006). SpeciÞcally, barkbeetles of the genus Dendroctonus Erichson have awide geographic distribution in North and CentralAmerica. They have caused signiÞcant economiclosses and irreversible ecological changes in Mexicanforests (Cibrian et al. 1995).

The development of Dendroctonus species usuallytakes place on mature trees (�15 yr) of the generaLarixMill.,PiceaA. Dietrich,PinusL., andPseudotsugaCarr. (Wood 1982). The advanced age classes of thesetrees constitute an adequate resource for bark beetlenourishment and reproduction. The sole exception tothis isDendroctonus rhizophagusThomas and Bright, aspecies that is endemic to the Sierra Madre Occidental(SMOC) mountain range in northwestern Mexico.D.rhizophagus colonizes and kills the seedlings andyoung saplings (�10 yr) of various pine species, es-pecially Apache pine (Pinus engelmannii Carr.), Du-rango pine (Pinus durangensis Martinez), Chihuahuapine (Pinus leiophylla Schlecht and Cham.), and Ar-izona pine (Pinus arizonica Engelm.) (Cibrian et al.1995, Salinas-Moreno et al. 2004, Sanchez-Martõnezand Wagner 2009). This unique bark beetle is a majorthreat to regeneration in pine forests throughout its

distribution in Mexico (Sanchez-Martõnez and Wag-ner 2009).

Although D. rhizophagus infests a signiÞcant num-ber of seedlings and young saplings every year, itsbiology, ecology, and geographic distribution arepoorly known (Estrada-Murrieta 1983, Salinas-Moreno et al. 2004, Sanchez-Martõnez and Wagner2009). Thus, analyzing its geographic distribution isessential to identify the environmental variables as-sociated with its presence in northwestern Mexico andto evaluate its potential threats in other geographicalareas of Mexico. Furthermore, if we assume that on-going climate change could affect the frequency andintensity of bark beetle outbreaks and the geographicdistribution of these insects and their hosts (Loganand Powell 2001, Kurz et al. 2008, Waring et al. 2009),there is a possibility thatD. rhizophagus could migratenorth into forests of the southwestern United Statesand cause widespread tree mortality in these largecontiguous forest. Thus, the purpose of this study is tomodel the potential distribution of D. rhizophagus inthe Sierra Madre Occidental of northwestern Mexicoby using BIOCLIM, ENFA, and MaxEnt to describethe bioclimatic conditions where it occurs and to iden-tify variables that determine its presence.

Materials and Methods

Biological Data.Collection records (museum spec-imens) for D. rhizophagus were obtained from themajor entomological collections in Mexico and fromtechnical reports of the Comision Nacional Forestal(CONAFOR) and the Secretarõa de Medio Ambientey Recursos Naturales (SEMARNAT) for the states ofChihuahua, Durango, Sonora, and Sinaloa (Table 1).In addition, we included records gathered in thecourse of several Þeld studies conducted from 2000 to2008. Although some areas of the SMOC morphotec-tonic province have been explored little, these recordsrepresent a reliable sample of the geographical distri-bution ofD. rhizophagus. A morphotectonic provinceis a zone having distinctive enough geomorphic andgeologic or tectonic features to differentiate it fromneighboring zones (Ferrusquõa-Villafranca 1998).

A preliminary database of 669 records was assem-bled to determine the altitudinal range where D. rhi-zophagus has been found most frequently and thefrequency of its incidence (IP) on different hosts. TheIP is a measure of the degree of occurrence of the barkbeetle on its host plants (Salinas-Moreno et al. 2004).All location records were georeferenced on 1:50,000topographic maps obtained from the INEGI. For thepotential distribution analysis, 412 records were elim-inated from the database, because they differed onlyminimally in geographic coordinates and elevation.Thus, the spatial modeling database was composed of257 presence-only records.

Similarly, a database of the principal hosts of D.rhizophagus was assembled from specimens storedin the national herbariums of Mexico and onlinerecords from the webpage of the Comision Nacionalpara el Conocimiento y uso de la de Biodiversidad

550 ENVIRONMENTAL ENTOMOLOGY Vol. 40, no. 3

(CONABIO). The Þnal database included 294 pres-ence-only records of 11 pine species. These recordswere georeferenced on 1:50,000 INEGI topographicmaps.

Study Area. The morphotectonic province of theSierra Madre Occidental (SMOC) is located between20� 30� and 31� 20� N and between 102� 20� and 109� 40�W. It covers 289,000 km2 and runs northwest to south-east. Elevations in this mountain chain range from200 m to slightly above 3,000 m. The climate is tem-perate, and the vegetation type is primarily pine forestand, to a lesser extent, oak and pine-oak forest. Themean annual precipitation ranges from 400 to 1,600mm (Rzedowsky 1978, Ferrusquõa-Villafranca 1998).Potential Geographic Distribution of D. rhizopha-gus by using BIOCLIM. A bioclimatic proÞle of D.rhizophagus was obtained from BIOCLIM in theDIVA-GIS v.5.2 software package (Hijmans et al.2002) by using 19 default temperature and precipitationvariables included in BIOCLIM (Table 2). This proÞledescribes the environment where the bark beetle hasbeenrecorded,andit isusedto identifyothersiteswherethe species may reside. The suitability limits of the spe-cies are characterized by the mean, standard deviation,minimum and maximum tolerance, and percentiles ofeach variable to a resolution of 2.5-arc min.

A principal component analysis (PCA) of theseclimatic variables was performed with STATISTICSv.7.0 toassess the relativecontributionofeachvariableto the bioclimatic proÞle of the species. In addition,histograms were developed to determine the statisti-cal distribution of these climatic variables. Variablesthat follow a normal distribution or a skewed distri-bution exert greater inßuences on the bioclimatic pro-Þle (Beaumont et al. 2005). Thus, those variables with-out a clearly deÞned distribution or a normaldistribution with a truncated histogram were elimi-nated. Subsequently, grid cells that included the bio-climatic proÞle ofD. rhizophaguswere identiÞed usingBIOCLIM. Grid cells were grouped into four catego-ries: null (areas outside the 0Ð100 percentile limits),low (areas within the 0Ð2.5 percentile limits), mod-erate to high (areas within 2.5Ð10 percentile limits),and very high to excellent (areas within the 10Ð100percentile limits). The scatter of cells associated with

Table 1. Collection records and reference sources of D.rhizophagus

Reference source CODENo. ofrecords

Comision Nacional Forestal,Chihuahua, Mexico

CONAFOR-CHIH 236

Comision Nacional Forestal,Sonora, Mexico

CONAFOR-SON 1

Comision Nacional Forestal,Sinaloa, Mexico

CONAFOR-SIN 2

Secretarõa de Medio Ambiente yRecursos Naturales, Durango,Mexico

SEMARNAT-DGO 119

Secretarõa de Medio Ambiente yRecursos Naturales de Sonora,Mexico

SEMARNAT-SON 24

Sanidad Forestal, Secretarõa delMedioambiente y RecursosNaturales, Mexico

SF-SEMARNAT 4

Secretarõa de RecursosNaturales y Medio Ambiente,Durango, Mexico

SRNyMA-DGO 31

Division de Ciencias Forestales,Universidad AutonomaChapingo

DCF-UACH 15

Instituto de Silvicultura,Universidad Autonoma deNuevo Leon.

IS-UANL 2

Coleccion Entomologica,Escuela Nacional de CienciasBiologicas-IPN

ENCB-IPN 14

Lab. de Variacion Biologica yEvolucion, Escuela Nacionalde Ciencias Biologicas-IPN

LVBE-ENCB 21

Unidad de Conservacion deEcosistemas Forestales,Madera, Chihuahua

UCODEFO-CHIH 142

Canadian National Collection ofInsects, Arachnids andNematodes-Ottawa

CNCI-Ottawa 10

Bibliographical data (Various authors) 48TOTAL 669

Table 2. Bioclimatic profile of D. rhizophagus for each location (obtained using BIOCLIM)

No. and climatic variable Min. Mean Max. SD 5% 10% 50% 90% 95%

1) Annual mean temp (�C) 10.2 12.8 19.2 1.4 11.1 11.3 12.6 14.6 15.52) Mean diurnal range (�C) 12.4 16.7 19.1 1.9 13.1 13.6 17.4 18.4 18.73) Isothermality 48.4 57.0 63.2 3.6 50.8 51.7 57.7 61.6 62.94) Temperature seasonality 306.9 472.6 671.9 101.3 319.9 337.0 486.1 601.1 605.95) Maximum temp of warmest month (�C) 21.4 27.5 35.6 2.4 23.2 24.1 27.7 30.3 30.86) Minimum temp of coldest month (�C) �5.8 �2.1 7.0 2.6 �5.2 �5.0 �2.6 1.5 2.37) Temperature annual range 21.0 29.5 35.5 4.2 21.9 22.9 31.1 33.8 34.28) Mean temp of wettest quarter (�C) 13.9 17.8 25.0 1.6 15.5 15.9 17.6 19.9 20.99) Mean temp of driest quarter (�C) 9.4 12.1 18.9 1.6 10.2 10.4 11.8 14.0 14.9

10) Mean temp of warmest quarter (�C) 14.4 18.4 25.8 1.7 16.0 16.5 18.3 20.6 21.711) Mean temp of coldest quarter (�C) 3.2 7.0 13.9 2.0 4.0 4.5 6.9 9.2 9.912) Annual precipitation (mm) 305.0 814.2 1,406.0 195.1 518.0 589.0 802.0 1,101.0 1,229.013) Precipitation of wettest month (mm) 79.0 198.1 321.0 44.6 131.0 152.0 189.0 267.0 278.014) Precipitation of driest month (mm) 2.0 10.9 18.0 3.2 6.0 7.0 11.0 15.0 15.015) Precipitation seasonality (mm) 78.4 93.8 126.2 9.5 80.0 82.2 93.1 108.7 112.216) Precipitation of wettest quarter (mm) 197.0 490.0 823.0 118.5 332.0 363.0 479.0 675.0 710.017) Precipitation of driest quarter (mm) 10.0 47.8 71.0 11.5 27.0 33.0 48.0 64.0 67.018) Precipitation of warmest quarter (mm) 170.0 440.4 746.0 110.6 276.0 325.0 437.0 614.0 628.019) Precipitation of coldest quarter (mm) 22.0 124.0 250.0 41.6 57.0 70.0 120.0 180.0 197.0

June 2011 MENDOZA ET AL.: FACTORS INFLUENCING THE GEOGRAPHICAL DISTRIBUTION OF Dendroctonus rhizophagus 551

these categories represents the potential distributionof the species.Potential Geographic Distribution of D. rhizopha-gus by using ENFA.Maps showing the major climaticvariables (determined by PCA), minimum annualtemperature, maximum annual temperature, and ele-vation were plotted with DIVA-GIS v.5.2. In addition,BIOCLIM was used to draw a map of the potentialdistribution of the hosts of D. rhizophagus to incor-porate this variable into the analysis.

All maps were imported into ArcView v.3.2 (Envi-ronmental Systems Research Institute 1999) and weretransformed to the Idrisi format. Subsequent analyseswere conducted with the Ecological Niche FactorAnalysis (ENFA) algorithm available in BIOMAPPERv.3.2 (Hirzel et al. 2006a). ENFA conducts a PCA ofecogeographic variables and generates a series of in-ternally uncorrelated factors that are used to plotpotential habitat or habitat suitability (HS) maps.Only signiÞcant factors were considered for the HSmap. Biologically signiÞcant factors included margin-ality and tolerance. Marginality is deÞned as the dif-ference between the optimum environment for thespecies and the global mean of environmental vari-ables for the study area. Tolerance indicates the de-gree of specialization of the species in relation to therange of available environments (Hirzel et al. 2002,2004). A habitat suitability score (0Ð100) was derivedfor each grid cell by comparing its marginality andspecialization coefÞcients with the median of eachfactor estimated for the study area. Specialization co-efÞcients are deÞned as the ratio of the ecologicalvariance in mean habitat to that observed for the focalspecies (Hirzel et al. 2002). HS scores were normal-ized into four categories: null to low (0Ð25), moderate(26Ð50), high (51Ð75), and very high to excellent(76Ð100).

The robustness and predictive ability of the HSmodel were evaluated in BIOMAPPER by a jackknife-type cross-validation by using k � 4 partitions of therecords of the species based on HubertyÕs rule, whichis a heuristic method (“rule of thumb”) that deter-mines the ratio of calibration and validation points(Fielding and Bell 1997). Three of the generated par-titions were used to calibrate the HS map, and thefourth was used to evaluate the outcome. The processwas replicated four times, and each group was used tovalidate the model for each replicate by using BoyceÕscontinuous index. This index measures the relation-ship between the expected and observed values fordifferent HS scores (Hirzel et al. 2006b). A value nearone indicates congruence between the expected andobserved values.Potential Geographic Distribution of D. rhizopha-gus by using MaxEnt. Maps of the major climaticvariables obtained from DIVA-GIS v.5.2, determinedwith the aid of PCA and histograms, plus maps ofminimum annual temperature, maximum annual tem-perature, elevation and host distribution, were ana-lyzed with MaxEnt v.3.3.1 (Phillips et al. 2004). TheMaxEnt algorithm was performed using the defaultparameters (500 iterations with a convergence thresh-

old of 10�5). During model development, 75% of thelocalities were used for model training, whereas 25%of the localities were held back to test the model.Suitable regularizationvalueswere included to reduceover-Þtting and were selected automatically by theprogram. The output format was the logistic defaultoption (Phillips and Dudõk 2008). The model wasevaluated by means of the area under the receiveroperating characteristic curve (AUC). AUCs near 0.5are similar to random predictions, which indicate aninaccurate model; in contrast, values above 0.9 indi-cate a highly accurate model (Swets 1988, Phillips andDudõk 2008). The resultant map showed the proba-bility (0Ð1) of species presence. Probability valueswere normalized into four categories: null to low (0Ð0.22), moderate (0.23Ð0.44), high (0.45Ð0.67), andvery high to excellent (0.68Ð1.0).

Results

The preferential altitudinal range forD. rhizophagusvaried between 2,000 and 2,600 m. The percentage ofincidence showed a high frequency of attacks by thisbeetle on P. arizonica, P. durangensis, P. engelmannii,and P. leiophylla (Table 3).Bioclimatic Profile of D. rhizophagus. The proÞle

based on 257 locations (Table 2) suggests that this barkbeetle may be present in sites supporting the followingtemperatures: annual mean temperatures of 10Ð19�C,maximum temperatures of the warmest month of 21Ð36�C, mean temperatures of the driest quarter of9Ð19�C, and mean temperatures of the coldest quarterof 3Ð14�C. In terms of precipitation, suitable sites arecharacterized by annual rainfalls of 305Ð1,406 mm,wettest-month rainfalls of 79Ð321 mm, driest-monthrainfalls of 2Ð18 mm, wettest-quarter rainfalls of 197Ð823 mm, and warmest-quarter rainfalls of 170Ð746 mm.

PCA showed that the Þrst three principal compo-nents accounted for 86.79% of the total variance (Ta-ble 4). Of the 19 variables that were analyzed, 13 hadhigh eigenvalues. However, of these 13 variables, onlyprecipitation variables 12, 13, 14, 16, and 18 and tem-perature variables 1, 5, 9, and 11 were normally dis-tributed. Temperature variables 2 and 7 had normalbut truncated distributions, whereas 3 and 4 did not Þt

Table 3. Incidence (%) of D rhizophagus on Pinus species inthe Sierra Madre Occidental, Mexico

Pinus species Incidence (%)

P. engelmannii 35.42P. durangensis 16.60P. arizonica 16.14P. leiophylla 8.67P. teocote 3.59P. herrerae 2.84P. lumholtzii 1.04P. strobiformis 0.75P. oocarpa 0.45P. cembroides 0.15P. ponderosa 0.15Pinus sp. 14.20Total no. of records 669

552 ENVIRONMENTAL ENTOMOLOGY Vol. 40, no. 3

any distribution. These last four variables were notused to generate potential distribution models withBIOCLIM, ENFA, or MaxEnt.Potential Geographic Distribution of D. rhizopha-gus by using BIOCLIM, ENFA, and MaxEnt. Thepotential distribution map obtained from BIOCLIMshowed a continuous area of moderate to high climatesuitability that encompassed the entire length of theSMOC morphotectonic province, as well as a nar-rower, continuous area with very high to excellentvalues (Fig. 1). The ENFA map showed a potentialdistribution that was similar to the distribution derivedby BIOCLIM but it had better deÞnition (Fig. 2). Thehigh and very high to excellent HS areas on the ENFAmap appeared as small, discontinuous patches scat-tered along the entire length of the SMOC morpho-tectonic province. Areas of moderate suitability oc-curred in the gaps between these patches. The mapgenerated with MaxEnt (Fig. 3) showed two separate,large areas of very high to excellent probability ofoccurrence in the northern part of the SMOC andsome small, discontinuous areas in its central and

southern regions. Moderate and high probability areaswere found in almost all the SMOC.

The global marginality factor was 2.76, suggestingthatD. rhizophagus is present in areas supporting bio-climatic conditions that differ from the average con-ditions in the SMOC. Global tolerance was 0.17, indi-cating that this species has a relatively narrowecological niche. The Þrst three factors explained93.4% of the total variation (Table 5). The marginalityfactor (number 1) explained only 4.7%, and the Þrsttwo tolerance factors (numbers 2 and 3) explained81.4% and 7.3%, respectively.

In terms of the marginality factor (Table 5), D. rhi-zophagus is essentially linked to the precipitation of thedriestmonth(0.352)andthefollowingtemperaturevari-ables: annual maximum temperature (�0.316), maxi-mum temperature of the warmest month (�0.340),mean temperature of the coldest quarter (�0.300), an-nual mean temperature (�0.300), and annual minimumtemperature(0.269).Inaddition,D.rhizophagus is linkedto host distribution (0.304) and elevation (0.276). Tol-erance factors indicate thatD. rhizophagus is associatedwith annual mean temperature (�0.707), mean temper-ature of the coldest quarter (�0.707), annual minimumtemperature (0.552), and annual maximum temperature(0.461). Jackknife-type cross-validation and evaluationbyusingBoyceÕscontinuous indexyieldedameanof0.67and a standard deviation of 0.32, both of which indicatedthe model is adequately robust.

MaxEnt analysis showed that the presence of D.rhizophagus is determined mainly by annual meantemperature (30.1%), elevation (13.4%), maximumannual temperature (12.8%), precipitation of thewarmest quarter (10.5%), precipitation of the driestmonth (9.8%), maximum temperature of the warmestmonth (8.1%), and precipitation of the wettest month(5.1%) (Table 6). The AUC of the training data was0.976 (�0.006), and the AUC of test points was 0.970(�0.006), indicating the model is robust.

Discussion

The spatial modeling of species distributions com-bining occurrence data with biotic, climatic, and top-ographic variables can be used to determine the bio-climatic proÞle of a species and, as a consequence, thepotential ranges in which it may reside (Anderson etal. 2003, Beaumont et al. 2005). This information pro-vides an initial explanation for the presence of a spe-cies in one area but not in another (Lindenmayer etal. 1991). For herbivorous insects, the host is an ex-tremely important limiting factor in terms of geo-graphical distribution, as insects cannot continue tooccur in those places where their host plants are notpresent (Spellerberg and Sawyer 1999, Lomolino et al.2006).

Although various studies suggest that the geo-graphic distribution of bark beetles at the macro-scalelevel is determined by variables such as host availabil-ity, temperature, and elevation (Amman 1973, Le-kander et al. 1977, Regniere and Logan 1996), fewstudies have examined the relative importance of

Table 4. Principal component analysis of BIOCLIM climaticvariables in relation to the occurrence of D. rhizophagus

Eigenvalues

Climatic VariablesComp.

1Comp.

2Comp.

3

1)a Annual mean temp �0.02 0.97 �0.182) Mean diurnal range

�Mean of monthly(max temp-mintemp)

0.71 �0.15 �0.44

3) Isothermality �0.74 0.28 0.074) Temperature seasonality 0.91 �0.23 �0.285)a Maximum temp of

warmest month0.83 0.35 �0.42

6) Minimum temp ofcoldest month

�0.67 0.66 0.20

7) Temperature annualrange

0.89 �0.21 �0.37

8) Mean temp of wettestquarter

0.66 0.62 �0.41

9)a Mean temp of driestquarter

�0.00 0.83 �0.41

10) Mean temp of warmestquarter

0.70 0.60 �0.36

11)a Mean temp of coldestquarter

�0.54 0.81 0.02

12)a Annual precipitation �0.94 �0.01 �0.2913)a Precipitation of wettest

month�0.81 0.11 �0.43

14)a Precipitation of driestmonth

�0.27 �0.56 �0.62

15) Precipitation seasonality 0.15 0.62 0.1716)a Precipitation of wettest

quarter�0.92 0.14 �0.24

17) Precipitation of driestquarter

�0.62 �0.52 �0.53

18)a Precipitation of warmestquarter

�0.88 0.14 �0.28

19) Precipitation of coldestquarter

�0.70 �0.30 �0.49

% Explainedvariance

48.18 25.58 13.03

aMost relevant variables according to histograms analysis.

June 2011 MENDOZA ET AL.: FACTORS INFLUENCING THE GEOGRAPHICAL DISTRIBUTION OF Dendroctonus rhizophagus 553

these variables (Ungerer et al. 1999, Carroll et al.2004). With respect to the host, D. rhizophagus has anarrow diet breadth, because it uses a small number(�40%) of the available pine hosts across its distribu-tion (Kelley and Farrell 1998). However, the results ofour study indicate that D. rhizophagus attacks almostall (�80%) species of pine present in its known dis-tribution which is limited to speciÞc areas within theSMOC, especially P. arizonica, P. engelmannii, P. leio-phylla, and P. durangensis (Wood 1982, Salinas-Moreno et al. 2010). The major incidence percentages(IPs) on these four pine species do not appear to berelated to the abundance of them in the SMOC, as allof the hosts used by this bark beetle are similarlyabundant elements in the pine forests of this morpho-tectonic province (Rzedowsky 1978, Perry 1991,

Garcõa Arevalo and Gonzalez Elizondo 2003, Ortega-Rosas et al. 2008). However, the observed IP valuescould be related to the ecologically heterogeneous orpatch distribution that these species have in theSMOC morphotectonic province. For example, whilePinus herrerae and P. leiophylla inhabit slopes andravines, P. arizonica and P. durangensis grow on pla-teaus and mesas.

However, our study also suggests that the distribu-tion of D. rhizophagus may not be limited by thegeographical distributions of its hosts (Table 6), asthere are areas in the SMOC morphotectonic provincewhere hosts are present but where there are no re-cords of the beetles presence. In addition, the geo-graphical distributions of the majority of them(�90%) extend over much of the range of this bark

Fig. 1. Potential distribution of D. rhizophagus in the Sierra Madre Occidental, Mexico modeled with BIOCLIM.

554 ENVIRONMENTAL ENTOMOLOGY Vol. 40, no. 3

beetle; for example, P. cembroides Zucc., P. herreraeMartõnez, P. leiophylla, P. lumholtzii Rob. & Fernald,P. oocarpa Schiede, and P. teocote Schiede ex Schltdl.& Cham. are present in other morphotectonic prov-inces in Mexico, and others such as P. cembroides, P.engelmannii, P. leiophylla, P. ponderosa Dougl., and P.strobiformis Engelm. are located beyond the Mexicanborder and occur in the southwestern United States(Farjon and Styles 1997). Differences in the geo-graphic distributions of D. rhizophagus and its hostsmight be the result of their natural history, where thespeciÞc biotic interactions and particular ecologicalrequirements of each of them have played importantroles.

Based on these results and those derived with ENFAand MaxEnt (Tables 5, 6), which suggest that the host

distribution has little weight as a predictor variable, wehypothesize that at the macro-scale level the potentialdistribution of D. rhizophagus is limited more by en-vironmental variables such as precipitation and tem-perature than by the availability or distribution ofhosts (Lekander et al.1977). In this context, thetolerance and marginality coefÞcients derived fromENFA suggest that this beetle has a relatively nar-row ecological niche with respect to some temper-ature and precipitation variables, inhabiting areaswith climatic conditions that differ from those gen-erally prevalent in the SMOC morphotectonic prov-ince. Areas of higher suitability may therefore besmaller and may appear in patches along the SMOC,as predicted by ENFA, compared with those pre-dicted by BIOCLIM.

Fig. 2. Habitat suitability (HS) for D. rhizophagus in the Sierra Madre Occidental, Mexico modeled with ENFA.

June 2011 MENDOZA ET AL.: FACTORS INFLUENCING THE GEOGRAPHICAL DISTRIBUTION OF Dendroctonus rhizophagus 555

The inßuence of elevation on the distribution of thisbeetle is not clear: although MaxEnt (Table 6) sug-gests that this factor is important, ENFA (Table 5)suggests the contrary. The relationship between tem-perature and elevation could be masking the effect ofthis factor on the altitudinal distribution of D. rhi-zophagus. The broad elevation range of this species(1,000Ð2,800 m), matches with the altitudinal range inwhich pine and pine-oak communities are found in theSMOC. In addition, the preferential elevation range(2,000Ð2,600 m) of D. rhizophagus is correlated withthe preferential altitudinal range of its main hosts(Farjon and Styles 1997, Lammertink et al. 1997) andrepresents areas identiÞed by BIOCLIM, ENFA, andMaxEnt as high to excellent habitats for this barkbeetle.

The actual distribution limits ofD. rhizophagus intothe north and south of the SMOC coincides with theisotherms for maximum temperature occurring be-tween 30 and 23� N of its distribution range. SpeciÞ-cally, these isotherms are annual temperatures of20Ð25�C and temperatures of the warmest month of23Ð30�C. Other bark beetle studies conducted at themacro-scale level report similar associations with cli-matic variables. For example, the distribution of themountain pine beetle,Dendroctonus ponderosaeHop-kins, in western Canada is limited by minimum tem-peratures, rather than by the availability of its pre-ferred hosts (ponderosa pine, P. ponderosa), despitetheir wide geographic distribution (Logan and Powell2001). Therefore, its distribution is limited in areasfurther north (Yukon and Northwest Territories) and

Fig. 3. Potential distribution of D. rhizophagus in the Sierra Madre Occidental, Mexico modeled with MaxEnt.

556 ENVIRONMENTAL ENTOMOLOGY Vol. 40, no. 3

east (Alberta) in Canada (Carroll et al. 2004). Ungereret al. (1999) reported that the isotherm for minimumannual temperature (�16�C) limits the dispersal ofD.frontalis Zimm. into more northerly areas of theUnited States.

Our study suggests that D. rhizophagus could po-tentially be distributed over a greater extent of theSMOC morphotectonic province. This will occur as aresult of both the use of a wide range of pine hostspresent in this province and the presence of numerousareas in the SMOC where optimal conditions of ele-vation, temperature, and humidity exist. Moreover,our results suggest that limits of the D. rhizophagusdistribution at the macro-scale level could be relatedto the maximum temperature isotherms and that op-timal temperature ranges where the bark beetle oc-curs are related to temperate habitats. Finally, its rel-atively narrow ecological niche with respect to sometemperature and precipitation variables and the pref-erence for temperate habitats, lead us to hypothesize

that even minor changes in climate may have signif-icant effects on the distribution and abundance of thisbark beetle. The apparent absence of this species frompotentially suitable habitats within the SMOC mor-photectonic province and the wide distribution of itshosts outside this province suggest that variations intemperature and humidity, such as those predicted totake place as a result of climate change (Logan andPowell 2001, Bale et al. 2002), may signiÞcantly affectits geographic distribution. A shift in the maximumtemperature isotherms may favor D. rhizophagus dis-persal into temperate areas or beyond its current highaltitudinal limit. More detailed modeling studies arerequired to determine the effects that climate changemay have on the distribution ofD. rhizophagus. It alsowould be important to explore how temperature andhumidity variables directly affect the development ofthe life cycle and population dynamics of this barkbeetle, given that our study suggests these variablescould also have strong inßuences at the meso- andmicro-scale levels, as has been demonstrated in otherDendroctonus species (Logan and Amman 1986, Bentzet al. 1991, Turchin et al. 1991).

Acknowledgments

We are grateful to Jose Luis Benito Rosas Ortiz(SEMARNAT-Sonora, Mexico); Ricardo Adan PeraltaDuran and Pedro Hernandez Dõaz (SEMARNAT-Durango,Mexico); Sergio Quinonez Barraza (CONAFOR Durango-Sinaloa, Mexico); Oscar de Leon Lara (CONAFOR-Sonora,Mexico); Marcos Daniel Trujano Thome and Jose Luis Agui-lar Vitela (Secretarõa de Recursos Naturales y Medio Ambi-ente, Durango, Mexico); and Guillermo Sanchez-Martõnez(INIFAP-Aguascalientes, Mexico) for providing us with ac-cess to their collection records and for their logistic support,time, and availability for consultations. We thank Jorge E.Macõas Samano (Colegio de la Frontera Sur, Chiapas,Mexico); Jane L. Hayes (PaciÞc Northwest Research Station,USDA Forest Service, LaGrande, OR); and three anonymousreviewers for their critical review of the manuscript. Theproject was funded by Comision Nacional Forestal(CONAFOR, 69539) and Secretarõa de Investigacion y Pos-grado-IPN (SIP-20090576). This work was part of M.G.M.sPh.D. dissertation. She was a Consejo Nacional de Ciencia yTecnologõa (207124) and Programa Institucional de For-macion de Investigadores del Instituto Politecnico Nacional(PIFI-IPN) fellow.

References Cited

Amman, G. D. 1973. Population changes of the mountainpine beetle in relation to elevation. Environ. Entomol. 2:541Ð547.

Anderson, R. P., D. Lew, and A. T. Peterson. 2003. Evalu-ating predictive models of speciesÕ distributions criteriafor selecting optimal models. Ecol. Modell. 162: 211Ð232.

Bale, S. J., G. J. Masters, I. D. Hodkinson, C. Awmack, T. M.Bezemer, V. K. Brown, J. Butterfield, A. Buse, J. C. Coul-son, J. Farrar, et al. 2002. Herbivory in global climatechange research: direct effects of rising temperature oninsect herbivores. Global Change Biol. 8: 1Ð16.

Beaumont, L. J., L. Hughes, and M. Poulsen. 2005. Predict-ing species distributions: use of climatic parameters in

Table 5. Percentage of variation explained by ecogeographicvariables for D. rhizophagus by using ENFA

Ecogeographicvariable

Factor 1 Factor 2 Factor 3

Marginality Specialization Specialization

(4.7%) (81.4%) (7.3%)

Annual mean temp �0.300 �0.707 �0.471Maximum temp of

warmest month�0.340 0.000 0.000

Mean temp of driestquarter

�0.165 0.000 0.000

Mean temp ofcoldest quarter

�0.300 �0.707 �0.511

Annual precipitation 0.249 0.000 0.000Precipitation of

wettest month0.216 0.000 0.000

Precipitation ofdriest month

0.352 0.000 0.000

Precipitation ofwettest quarter

0.202 0.000 0.000

Precipitation ofwarmest quarter

0.249 0.000 0.000

Minimum annualtemp

�0.269 0.000 0.552

Maximum annualtemp

�0.316 0.000 0.461

Elevation 0.276 0.000 0.000Pines distribution 0.304 0.000 0.000

Table 6. Percentage of estimated contribution for ecogeo-graphic variables by using MaxEnt

Variable % Contribution

Annual mean temp 30.1Elevation 13.4Maximum annual temp 12.8Precipitation of warmest quarter 10.5Precipitation of driest month 9.8Maximum temp of warmest month 8.1Precipitation of wettest month 5.1Minimum annual temp 3.2Pines distribution 2.5Mean temp of driest quarter 1.5Precipitation of wettest quarter 1.2Mean temp of coldest quarter 1.1Annual precipitation 0.6

June 2011 MENDOZA ET AL.: FACTORS INFLUENCING THE GEOGRAPHICAL DISTRIBUTION OF Dendroctonus rhizophagus 557

BIOCLIM and its impact on predictions of species cur-rent and future distributions. Ecol. Modell. 186: 250Ð269.

Bentz, B. J., J. A. Logan, and G. D. Amman. 1991. Temper-ature dependent development of the mountain pine bee-tle (Coleoptera: Scolytidae) and simulations of its phe-nology. Can. Entomol. 123: 1083Ð1094.

Braunisch, V., and R. Suchant. 2007. A model for evaluatingthe “habitat potential” of a landscape for capercaillieTetrao urogallus: a tool for conservation planning. Wildl.Biol. 13: 21Ð33.

Brown, J. H. 1984. On the relationship between abundanceand distribution of species. Am. Nat. 124: 255Ð279.

Brockerhoff, E. G., A.M. Liebhold, andH. Jactel. 2006. Theecology of forest insects invasions and advances in theirmanagement. Can. J. For. Res. 36: 263Ð268.

Carroll, A. L., S. W. Taylor, J. Regniere, and L. Safranyik.2004. Effects of climate change on range expansion bythe mountain pine beetle in British Columbia, pp. 223Ð232. In T. L. Shore, J. E. Brooks, and J. E. Stone (eds.),Mountain Pine Beetle Symposium: Challenges and Solu-tions. 30Ð31 October 2003, Kelowna, British Columbia,Canada, Natural Resources Canada, Canadian Forest Ser-vice, PaciÞc Forestry Centre, Information Report. BC-X-399, Victoria, BC.

Cibrian, T. D., J. T. Mendez M., R. Campos B., H. O. YatesIII, and J. FloresL. 1995. Insectos Forestales de Mexico/Forest Insects of Mexico. Universidad Autonoma Chap-ingo, Mexico.

[CONABIO] Comision Nacional para el Conocimiento yUso de la Biodiversidad. 1997. Provincias BiogeograÞcasde Mexico. Escala 1:4,000,000. Mexico.

Environmental Systems Research Institute. 1999. Arcview-Gis 3.2. Environmental Systems Research Institute, Red-lands, California.

Estrada-Murrieta, O. 1983. Biologõa del descortezador delrenuevo de pino Dendroctonus rhizophagus T. y B. (Col.:Scolytidae) en la region de la Mesa del Huracan, Chi-huahua. BachelorÕs Thesis. Universidad Autonoma Chap-ingo, Mexico.

Farjon, A., and B. Styles. 1997. Flora Neotropica. Mono-graph 75. Pinus (Pinaceae). Organization for Flora Neo-tropica. New York Botanical Garden, NY.

Ferrusquıa-Villafranca, I. 1998. Geologõa de Mexico: UnaSinopsis, pp. 3Ð108. InT. P. Ramamoorthy, R. Bye, A. Lot,and J. Fa (eds.), Diversidad Biologica de Mexico:Orõgenes y Distribucion. Instituto de Biologõa. Universi-dad Nacional Autonoma de Mexico, Mexico.

Fielding, A.H., and J. F. Bell. 1997. A review of methods forthe assessment of prediction errors in conservation pres-ence/absence models. Environ. Conserv. 24: 38Ð49.

Ganeshaiah, K. N., N. Barve, N. Nath, K. Chandrashekara, M.Swamy, and R. U. Shanker. 2003. Predicting the poten-tial geographical distribution of the sugarcane woollyaphid using GARP and DIVA-GIS. ScientiÞc Correspon-dence. Curr. Sci. 85: 1526Ð1528.

Garcıa Arevalo, A., and M. S. Gonzalez Elizondo. 2003.Pinaceas de Durango. 2a. ed. Instituto de Ecologõa, A. C.Mexico.

Gaston, J. K. 2003. The structure and dynamics of geo-graphic ranges. Oxford University Press, Oxford, NewYork.

Hijmans, R. J., L. Guarino, C. Bussink, and E. Rojas. 2002.DIVA-GIS, version 5.2. A geographic information systemfor the analysis of biodiversity data. Manual. InternationalPotato Center, Lima, Peru.

Hirzel, A., J. Hausser, D. Chessel, and N. Perrin. 2002. Eco-logical niche factor analysis: how to compute habitat

suitability maps without absence data? Ecology 83: 2027Ð2036.

Hirzel,A.,B.Posse,P-A.Oggier,Y.Crettenand,C.Glenz, andR. Arlettaz. 2004. Ecological requirements of reintro-duced species and the implications for release policy: thecase of the bearded vulture recolonizing the Alps. J. Appl.Ecol. 41: 1103Ð1116.

Hirzel, A. H., J. Hausser, and N. Perrin. 2006a. Biomapper3.2. Laboratory for Conservation Biology. Department ofEcology and Evolution. University of Lausanne, Laus-anne. (http://www.unil.ch/biomapper).

Hirzel, A.H., G. LeLay, V.Helfer, C. Randin, andA.Guisan.2006b. Evaluating the ability of habitat suitability modelsto predict species presences. Ecol. Modell. 199: 142Ð152.

Jimenez-Valverde, A., V. M. Ortuno, and J. M. Lobo. 2007.Exploring the distribution of Sterocorax Ortuno, 1990(Coleoptera, Carabidae) species in the Iberian Peninsula.J. Biogeogr. 34: 1426Ð1438.

Kelley, S. T., andB.D. Farrell. 1998. Is specialization a deadend? The phylogeny of host use in Dendroctonus barkbeetles (Scolytidae). Evolution 52: 1731Ð1743.

Kurz, W. A., C. C. Dymond, G. Stinson, G. J. Rampley, E. T.Neilson, A. L. Carroll, T. Ebata, and L. Safranyik. 2008.Mountain pine beetle and forest carbon feedback to cli-mate change. Nature 452: 987Ð990.

Kumar, S., and T. J. Stohlgren. 2009. Maxent modeling forpredicting suitable habitat for threatened and endan-gered tree Canacomyrica monticola in New Caledonia. J.Ecol. Nat. Environ. 1: 94Ð98.

Lammertink, J. M., J. A. Rojas-Tome, F. M. Casillas-Orona, yR.L.Otto. 1997. Situacionyconservacionde losbosquesantiguos de pino-encino de la Sierra Madre Occidental ysus aves endemicas. Consejo Internacional para la preser-vacion de las aves, seccion Mexicana (CIPAMEX), D. F.Mexico.

Lekander, B., B. Bejer-Petersen, B. Kangas, and A. Bakke.1977. The distribution of bark beetles in the NordicCountries. Acta Entomol. Fenn. 32: 1Ð37.

Lindenmayer, D. B., H. A. Nix, J. P. McMahon, M. F.Hutchinson, and M. T. Tanton. 1991. The conservationof LeadbeaterÕs possum,Gymnobelideus leadbeateri (Mc-Coy): a case study of the use of bioclimatic modelling.J. Biogeogr. 18: 371Ð383.

Logan, J. A., and G. D. Amman. 1986. A distribution modelfor egg development in mountain pine beetle. Can. En-tomol. 118: 361Ð372.

Logan, J. A., and J. A. Powell. 2001. Ghost forests, globalwarming and the mountain pine beetle (Coleoptera: Sco-lytidae). Am. Entomol. 47: 160Ð172.

Lomolino, M. V., B. R. Riddle, and J. H. Brown. 2006. Bio-geography. Sinauer Inc., Sunderland, Massachusetts.

MacDonald, G.M. 2003. Biogeography, space, time and life.Wiley. Inc., New York, NY.

Mackey, B. G., and D. B. Lindenmayer. 2001. Towards ahierarchical framework for modelling the spatial distri-bution of animals. J. Biogeogr. 28: 1147Ð1166.

Nix, H. 1986. A biogeographic analysis of Australian elapidsnakes, pp. 4Ð15. In R. Longmore (ed.), Atlas of elapidsnakes of Australia. Australian ßora and fauna series num-ber 7. Australian Government Publishing Service, Can-berra, Australia.

Ortega-Rosas, C. I., M. C. Penalba, J. A. Lopez-Saez, y T. R.Van Devender. 2008. Retrospectiva del Bosque de Pinoy Encino de la Sierra Madre Occidental, Sonora, Noroestede Mexico, Hace 1,000 anos. Acta Bot. Mex. 83: 69Ð92.

Pearson, R. G., and T. P. Dawson. 2003. Predicting the im-pacts of climate change on the distribution of species: are

558 ENVIRONMENTAL ENTOMOLOGY Vol. 40, no. 3

bioclimate envelope models useful? Glob. Ecol. Biogeogr.12: 361Ð371.

Perry, J. P., Jr. 1991. The Pines of Mexico and CentralAmerica. Timber Press Inc., Portland, Oregon.

Phillips, S. J., and M. Dudık. 2008. Modeling of species dis-tributions with MaxEnt: new extensions and a compre-hensive evaluation. Ecography 31: 161Ð175.

Phillips, S. J., R. E. Schapire, and M. Dudık. 2004. A maxi-mum entropy approach to species distribution modeling,pp. 655Ð662. InR. Greinerand and D. Schuurmans (eds.),Proceedings: The Twenty-First International Conferenceon Machine Learning, 4Ð8 July 2004, Banff, Canada. ACMPress, New York.

Regniere, J., and J. A. Logan. 1996. Landscape-wide projec-tion of temperature-driven processes for seasonal pestmanagement decision support: a generalized approach,pp. 43Ð56. In T. L. Shore and D. A. MacLean (eds.),Proceedings, Symposium: Decision Support Systems inForest Pest Management. Entomological Society of Can-ada, FRDA Report No. 260. Victoria, BC.

Rzedowsky, J. 1978. Vegetacion de Mexico. Limusa. D.F.Mexico.

Salinas-Moreno, Y., M. G. Mendoza, M. A. Barrios, R. Cis-neros, J. Macıas-Samano, and G. Zuniga. 2004. Areogra-phy of the genus Dendroctonus (Coleoptera: Curculion-idae: Scolytinae) in Mexico. J. Biogeogr. 31: 1163Ð1177.

Salinas-Moreno, Y., A. Ager, C. F. Vargas, J. L. Hayes, and G.Zuniga. 2010. Determining the vulnerability of Mexicanpine forest to bark beetles of the genus DendroctonusErichson (Coleoptera: Curculionidae: Scolytinae). For.Ecol. Manage. 260: 52Ð61.

Sanchez-Martınez, G., and M. R. Wagner. 2009. Host pref-erence and attack pattern of Dendroctonus rhizophagus(Coleoptera: Curculionidae: Scolytinae): a bark beetlespecialist on pine regeneration. Environ. Entomol. 38:1197Ð1204.

Sattler, T., F. Bontadina, A. H. Hirzel, and R. Arlettaz. 2007.Ecological niche modelling of two cryptic bat speciescalls for a reassessment of their conservation status.J. Appl. Ecol. 44: 1188Ð1199.

Spellerberg, I. F., and J.W.D. Sawyer. 1999. An introductionto applied biogeography. Cambridge University Press,Cambridge.

Swets, J. A. 1988. Measuring the accuracy of diagnostic sys-tems. Science 240: 1285Ð1293.

Thomas, C. D., E. J. Bodsworth, R. J. Wilson, A. D. Simons,Z. G. Davies, M. Musche, and L. Conradt. 2001. Ecolog-ical and evolutionary processes at expanding range mar-gins. Nature 411: 577Ð581.

Titeux, N., M. Dufrene, J. Radoux, A. H. Hirzel, and P.Defourny. 2007. Fitness-related parameters improveniche-based distribution modelling: the case of the red-backed shrike. Biol. Conserv. 138: 207Ð223.

Turchin, P., P. L. Lorio, Jr., A. D. Taylor, and R. F. Billings.1991. Why do populations of southern pine beetles (Co-leoptera: Scolytidae) ßuctuate? Environ. Entomol. 20:401Ð409.

Ungerer, M., M. P. Ayres, and M. J. Lombardero. 1999. Cli-mate and the northern distribution limits ofDendroctonusfrontalis Zimmermann (Coleoptera: Scolytidae). J. Bio-geogr. 26: 1133Ð1145.

Waring, K. M., D. M. Reboletti, L. A. Mork, M. Li, C. H.Huang,R.W.Hofstetter,A.M.Garcia,P.Z.Fule, andT.S.Davis. 2009. Modeling the impacts of two bark beetlespecies under warming climate in the southwesternU.S.A.: ecological and economic consequences. Environ.Manage. 44: 824Ð835.

Wood, S. L. 1982. The bark and ambrosia beetles of Northand Central America (Coleoptera: Scolytidae) a taxo-nomic monograph. Great Basin Nat. Mem. 6: 1Ð1359.

Received 2 March 2010; accepted 7 April 2011.

June 2011 MENDOZA ET AL.: FACTORS INFLUENCING THE GEOGRAPHICAL DISTRIBUTION OF Dendroctonus rhizophagus 559