Dispersal of the pea leaf minerRev. FCA UNCUYO. 2019. 51(2): 343-352. ISSN (en línea) 1853-8665.

Dispersal of the pea leaf miner huidobrensis (Blanchard, 1926) (Diptera: ): a field experiment

Dispersión del minador de hojas Liriomyza huidobrensis (Blanchard, 1926) (Diptera: Agromyzidae): un experimento a campo

María Silvina Fenoglio 1, Martín Videla 1, Adriana Salvo 1, Juan Manuel Morales 2

Originales: Recepción: 24/10/2017 - Aceptación: 09/05/2018

Abstract

Movement of herbivore within agroecosystems can ultimately determine the level of damage to crops. It was evaluated the dispersal of Liriomyza huidobrensis evaluated if body size was related to dispersal distance. Eight hundred individuals of L.(Blanchard, huidobrensis 1926) were (Diptera: released Agromyzidae)in the central point under of field a series conditions. of concentric In addition, circles (6, it was 12, 24, 48, and 96 m in radius), where 117 yellow sticky traps were set to recapture. applied to analyze dispersal probability in relation to distance. Five percent of the releasedCircular Statistics individuals were were used recaptured, to evaluate being flight 36 direction m the median whereas distance hurdle models of recapture. were The distribution of recaptured insects was not random around the circles, but the preferential dispersal direction was not explained by wind direction. The incidence of recaptured L. huidobrensis from the release point, but decayed faster for females. No effect was found of body size on the distance of recapture. of bothThe results sexes significantly suggest that decreased L. huidobrensis at increasing dispersed distances mainly overKeywords short distances with males being capable of performing longer flights than females.

body size • herbivore insect pest • movement • wind

1 Universidad Nacional de Córdoba (UNC)- CONICET. Instituto Multidisciplinario de Biología Vegetal (IMBIV)- Centro de Investigaciones Entomológicas de

[email protected] 2 UniversidadCórdoba- FCEFyN. Nacional Av. delVélez Comahue-CONICET. Sarsfield 1611. (X5016GCA). Laboratorio Córdoba. Ecotono. Argentina. INIBIOMA.

Pasaje Gutiérrez 1125. S. C. de Bariloche, (AR-8400). Río Negro. Argentina.

Tomo 51 • N° 2 • 2019 343 M. S. Fenoglio, M. Videla, A. Salvo, J. M. Morales

Resumen determinar el nivel de daño a los cultivos. Se evaluó experimentalmente la dispersión de ElLiriomyza movimiento huidobrensis de insectos (Blanchard, herbívoros 1926) dentro (Diptera: de agroecosistemas Agromyzidae) puede a campo finalmente y se analizó si el tamaño corporal se relacionó con la distancia de dispersión. Se liberaron ochocientos individuos en el punto central de una serie de círculos concéntricos (6, 12, 24, 48 y 96 m de radio), donde se colocaron 117 trampas pegajosas amarillas para recap- turar los insectos. Se utilizó estadística circular para evaluar la dirección del vuelo, y modelos "hurdle" para analizar la probabilidad de dispersión en relación a la distancia. Se recapturó el 5% de los individuos liberados siendo 36 m con la distancia mediana de recaptura. La distribución de insectos alrededor de los círculos no fue aleatoria, pero la dirección de dispersión preferencial no se explicó por la dirección del viento. La incidencia de L. huidobrensis la distancia de recaptura. Estos de ambos resultados sexos sugierendisminuyó que significativamente L. huidobrensis al se aumentar dispersó principalmentela distancia, pero a distanciasdecayó más cortas, rápidamente siendo los para machos hembras. capaces El tamaño de desarrollar corporal vuelos no influyó más largos que las hembras.

Palabras clave

tamaño corporal • insecto herbívoro plaga • movimiento • viento Introduction

Dispersal is a key process in ecology Differential resource allocation between affecting the dynamics and persistence of sexes contributes to sex-biased dispersal. insect populations (6). The term 'dispersal' In many insect species, female dispersal is can be used to describe both the movements limited due to investments of resources in towards a breeding location which implies egg production (3). Regarding body size, it is expected that larger individuals have a movements represented by short foraging higher active dispersal rate than smaller gene flow across space, and 'trivial' ones (10, 13). It has been shown that passive (wind) components can account relatively longer wings would increase flights (3). Both active (locomotory) and acceleration capacity and improve the relative contribution of each one varying amongfor the speciesdispersal (17, of 23).flying insects, with the dispersal by wind is likely to be important Active dispersal can be affected by forefficiency minute of insectsprolonged (8) flights but in(4). generalPassive both environmental factors and by the little is known about how the dispersal movement ability of individuals which in turn depends on certain factors such landscape (12, 20). as age, stage, sex and body size (3). of insects is influenced by wind in the

344 Revista de la Facultad de Ciencias Agrarias Dispersal of the pea leaf miner

In agro-ecosystems, the dispersal of Material and methods herbivores can ultimately determine its spatial distribution and the level of damage to crops. Thus, understanding the dispersal experiment was performed in order to process in insect pests might be helpful to evaluateA release movement and of recaptureL. huidobrensis field improve management strategies (15). individuals. The study was conducted in a The pea leaf miner, Liriomyza huido- brensis (Blanchard, 1926) (Diptera: Agromyzidae), is a polyphagous species Córdoba10-ha lawn city field (31°16'58.5" for commercial S, 64°02'31.04" purposes. that causes serious damage to several W),The field located was placed in Córdoba in the outskirts province of horticultural crops (2, 19). Originally from (Argentina), where no L. huidobrensis South America, this species has expanded host plants were found after conducting its range and invaded many regions of the world, frequently reaching pest status of 800 unfed 1-2 days old individuals of (26). Damage to plants is mainly caused by L.a careful huidobrensis inspection reared of onthe Vicia field. faba A total L. larvae which excavate tunnels consuming (broad bean) plants in the laboratory, mesophyll cells leaving intact the upper were released in the experiment in a and lower leaf surfaces. This species single release event. tends to be resistant to commonly used Field collected L. huidobrensis adults pesticides, thus different management were offered broad beans in the laboratory. strategies are applied to control it (25). Pots (28 cm 15 cm, 9 cm deep) containing For trap cropping, for example, it is 8-10 broad beans (6-10 leaves each) were relevant to know the dispersal behavior of placed in a cage (glass, wood and voile, the target herbivore in order to maximize 30 cm side length) with 20 pairs for 4 h, at pest control by adequately locating trap room temperature, for mating and ovipo- plants in relation to main crop (7). sition. This procedure was carried out Studying the movement of insects simultaneously in several cages to obtain in their natural habitat is essential a high number of adults. for understanding their biology and To recapture the insects, 117 yellow behavior but sometimes represents a plastic and opaque sticky traps were used (20 cm x 20 cm) which are known to be insects like adult dipteran leaf miners. One highly attractive for leaf miners (14). alternativegreat challenge, to study especially dispersal for smallpatterns flying in insects is the Eulerian approach which relies on physical or biological 'traps' at theTraps central were point placed (6, in 12, the 24, field 48, andin a 96 design m in some particular points where dispersers of concentric circles at five distances from are sampled (16). 1a, page 346). radius),The where location of were the released releasing (figure point patterns of Liriomyza species (5, 11, 18, 28), noneAlthough of them there tested are a dispersal few studies through on flight a of this position on this results is unlikely givenwas arbitrarily the spatial selected homogeneity but any of influencethe lawn The aim of the present study was twofold:release and i) torecapture evaluate field the experiment. mean direction every 9.5 to 10 m on each ring and placed and distance of movement of L. huidobrensis atfield. 50 Traps cm above were ground spaced (from out regularlythe base individuals, and ii) to examine if body size of the trap) as this is the height were was related to dispersal distance. L. huidobrensis usually disperses (14).

Tomo 51 • N° 2 • 2019 345 M. S. Fenoglio, M. Videla, A. Salvo, J. M. Morales

Figure 1. Liriomyza huidobrensis (a) individuals Diagram ofwere the released.field trial Differentindicating symbols the position indicate of the different 117 traps number set at of re-capturesfive distances (b) Circular (6, 12, 24, histogram 48, and 96illustrating m) from windthe central direction point and where speed during the experiment. Each ring represents the time proportion (%) of winds from different directions. The arrow indicates the prevailing wind direction. Figura 1. (a) Diagrama del ensayo a campo que indica la posición de las 117 trampas establecidas a cinco distancias (6, 12, 24, 48 y 96 m) a partir del punto central donde se liberaron los individuos de Liriomyza huidobrensis. Diferentes símbolos indican diferente número de re-capturas (b) Histograma circular que ilustra las direcciones y velocidad del viento durante el experimento. Cada anillo representa la proporción de

dirección predominante del viento. tiempo (%) en la que soplaron vientos desde diferentes direcciones. La flecha indica la

Leaf miners were released on July 16th to males). A meteorological station Nexus (TFA) placed nearby the experiment leaved the box where they were trans- automatically recorded wind speed and ported)2014 at and 11.30 traps am were (after removed 15 min 24 all h later flies direction every 20 minutes. and carried to the laboratory where they were inspected for L. huidobrensis adults. Data analysis The individuals collected were stored Flight direction of L. huidobrensis was in tubes with 70% ethanol to posterior analyzed by means of Circular Statistics. A Rayleigh test for Randomness was individual, it was determined its sex and performed to determine if the distri- bodyconfirmation size through of its the identity. measurement For each of bution of captures was distinguishable forewing length (mm). from random. To test whether the leaf The group of insects released had a miner re-capture direction was the same female biased sex ratio (1.77:1 females as the wind direction, a one-sample test,

346 Revista de la Facultad de Ciencias Agrarias Dispersal of the pea leaf miner analogous to a one-sample t-test for data on a linear scale, was conducted using sex since females of L. huidobrensis are wind direction as a pre-assigned angle generallyGeneral Lineallarger than Models males separately(24). by direction. Particularly, it was tested if to be tested against the leaf miner flight Results 180° to the mean wind direction. insectsThe angle flew (degrees)downwind of by each subtracting trap was Five percent of the released individuals measured in relation to the release point were recaptured in the traps. Although the (through a drawing program) and then liberated population was female biased transformed into radians, which is the unit there was a higher number of males that Circular Statistics use. The analyses recaptured (n=29) in relation to females were performed by using the package (n=14). Considering the occupied traps "circular" (1) in R 3.2.2 (21). (n=28), the mean distance of recapture To assess leaf miner dispersal in of L. huidobrensis was of 46.71 m and the relation to distance from release we median 36 m. analyzed data (N = 117) with hurdle The average wind direction during the experiment was 205° (i.e. direction from models (30) allowing to model the excess which wind blows, N = 0°, E = 90°, S = 180°, ofmodels zeros whichin the trap are arecaptured type of zero-inflated population W = 270°), which indicates that the winds were preponderantly from the Southwest were performed separately for males and direction, with a wind speed of 0.4 m/s (figure 1, page 346). These models, which at the release event and with a mean of part is a binary outcome or incidence 0.65 m/s (range 0-2.4 m/s) during the model,females, and have the two-parts second part whereby is a truncated the first count model (30). 1b, page 346). In this models the explanatory variable wholeThe duration distribution of the of experimentrecaptured insects(figure for both parts was distance from release (m), the response variable was the around the different circles according to(figure the 1 Rayleigh a, page test 346) (z=0.28, was not P random=0.03), the assumed distributions were binomial but their preferential direction was not fornumber the incidence of recaptured model flies and per Poisson trap, and for explained by wind direction (z=0.27, the count model. Using this approach, P=0.005). In fact, the mean direction of it is possible to estimate the probability insect recaptures was 188° which was of recapture of each trap through the upwind with respect to prevailing winds. incidence model and then given L. huido- Hurdle models indicated that there brensis individuals are present, estimate was no effect of distance on the number of the relative mean number of individuals recaptured individuals (count model) of through the count model. Parameter either sex (table 1, page 348). estimates for distance were tested for However, the probability of recaptured L. huidobrensis - models were performed using the package cantly decreased with distance for both psclsignificance (29) in R using 3.2.2 (21). a Z- test. The hurdle (incidence model) signifi Data on recapture distance were 348). related to individual body size through males and females (table 1; figure 2, page

Tomo 51 • N° 2 • 2019 347 M. S. Fenoglio, M. Videla, A. Salvo, J. M. Morales

Table 1. Hurdle models for females and males of Liriomyza huidobrensis relating the distance from the release point to the probability that the species is present in a trap (incidence model) and the relative mean number of individuals in a trap (count model). Tabla 1. Modelos "Hurdle" para hembras y machos de Liriomyza huidobrensis relacionando la distancia desde el punto de liberación y la probabilidad de que la especie esté presente en una trampa (modelo de incidencia) y el número medio relativo de individuos en una trampa (modelo de conteo).

Sex Parts of the hurdle model Estimate Z P Incidence model -0.042 -3.228 0.001 Females Count model -0.019 -0.518 0.605 Incidence model -0.015 -2.112 0.034 Males Count model 0.004 0.401 0.688

Figure 2. Relationship between the distance from the release point and the probability of the presence of Liriomyza huidobrensis females and males in each trap. There was a higher probability that males reached higher dispersal distances than females as it is shown by superimposed dots in the right panel. Figura 2. Relación entre la distancia desde el punto de liberación y la probabilidad de presencia de hembras y machos de Liriomyza huidobrensis en cada trampa. Existe una mayor probabilidad de que los machos alcancen distancias de dispersión mayores que hembras tal como lo muestran los puntos super impuestos del panel derecho.

Regarding body size, there was not a distance parameters of females (95% CI: Although -0.072, -0.019) confidence and males intervals (95% forCI: significant relationship between wing -0.03, -0.001) were superposed by 21%, either sex (Females: F1,12=0.001, P=0.97; length and dispersal distance of flies for the probability of recapture of each trap Males: F1,27=1.53, P=0.22). decayed faster with distance for females.

348 Revista de la Facultad de Ciencias Agrarias Dispersal of the pea leaf miner

Discussion

The results of this study show that In addition, the low variability in wing the pea leaf miner dispersed mainly over length among released individuals of both short distances with males being capable sexes (CV<10%) could explain that result. Future experiments using a larger number of insects and low and high quality host pointof performing usually leadlonger to flightsa decline than in females. density plants would provide individuals with distributionDispersal of flying at insects increasing from a distances specific higher size differences (24). (9, 22, 23). Although could not be found a decreasing pattern for the abundance of Agromyzid flies are considered to hostbe "moderate plants (28), fliers" but (27) at the because same time, they was found that the higher probability to theytend canto move very longer short distances distances between by wind flies L.recaptured huidobrensis in wasrelation near to to distance,the release it dispersal (27). The results presented here point, eventually falling with distance, showed that under the conditions of this andfind decaying faster for females. This result indicates that this sex was the most effect on L. huidobrensis dispersal sensitive to longer distances, were a few suggestingexperiment, that wind insects had not may a be significant in part individuals were re-captured. Increased controlling their long distance movements captures of pea leaf miner males at long (20). This is in agreement with the activity as they actively search for mates for L. trifolli, a closely related species to anddistances food may whereas be explained females by spend high moreflight L.findings huidobrensis by Jones. These and authors Parrella showed (1986), time on leaves for oviposition (19). that L. trifolli dispersed 26 m on average, Movement patterns could be affected but was capable to reach the furthest by behavioral, morphological, physio- distance tested (102 m) in a greenhouse logical, genetic, and developmental attri- experiment free of wind. butes of focal individuals, as well as by In this study, wind speed at the environmental variables (16). moment of insect release was really low The higher number of males recap- not exceeding 0.4 m/s and considering tured in relation to females, despite the that mean direction of insect displacement released group was female biased, could was not downwind, it is unlikely that be explained by differences in the level of attraction of the traps. dispersal of the pea leaf miner in the trial. Martin et al. (2005), showed that windHowever, speed as and it was direction one release did influence event it translucent yellow sticky traps were more is uncertain what would happen with the attractive for females than opaque traps, dispersal pattern at higher wind speeds. with no differences for males. Contrary to Aware that the models used to this expectations, it was could not detect analyze L. huidobrensis dispersal are an effect of the wing length of individuals phenomenological, and thus fail to with dispersal distance reached by the discriminate detailed components of the leaf miners of either sex, probably due to dispersal process (6), but the low rate the low number of recaptures. of recaptures did not allow performing mechanistic models with reliable results.

Tomo 51 • N° 2 • 2019 349 M. S. Fenoglio, M. Videla, A. Salvo, J. M. Morales

Nevertheless, these analyses allow All in all, the information obtained in this describing and predicting general patterns of dispersal for the pea leaf miner which development of proper control strategies - withinfield experiment an integrated could pesthelp managementto the future standing the dispersal process of this pest program and to develop forecasting represents the first step to begin under systems to alert farmers about pest potential of L. huidobrensis is relevant by invasions and outbreaks (17). differentin the field. reasons. Knowledge At local ofscale, the it dispersal could be useful in crop pest control, for the appli- cation of trap cropping strategy being the Conclusions knowledge on adult dispersal important to decide trap plants location (7). In general the pea leaf miner showed In addition, the use of yellow traps a low dispersal capacity in a homogenous as monitoring tool for the pea leaf miner landscape. Nevertheless, the potential could be improved by adjusting space consequences of this behavior should between traps (11). At landscape scale, be analyzed considering the generalist knowing mean dispersal distance of feeding habitat of the species. Future L. huidobrensis could be important to studies on the topic should incorporate predict its potential of spread to other landscape in order to have bigger picture On the other hand, considering that different spatial configurations of the L.fields huidobrensis (19). is invasive in numerous estimates, as those here showed is parts of the world (26), the knowledge essentialof the process. to understand Obtaining thefield ecological dispersal of dispersal parameters is relevant since and evolutionary bases of the dynamics they can have an overwhelming effect on and persistence of insect populations and invasion speed (15). At the same time, would help to guide a proper design of the dispersal information could serve to management policies of this worldwide model the future areas of colonization (in pest species. altitude and latitude) of the pest under scenarios of global climate change (15).

References

1. Agostinelli, C.; Lund, U. 2017. R package 'circular': Circular Statistics (version 0.4-93). Available in: https://r-forge.r-project.org/projects/circular/ (Accessed 25 July 2017).

Liriomyza Melia 2. Ávalos, D.azedarach S.; Ricobelli, L.: resultadosG.; Palacios, preliminares. S. M.; Defagó, Revista M. T. 2013. de la Evaluación Facultad de de Cienciasla preferencia Agrarias. de Universidad spp. Nacional en genotipos de Cuyo. diferentes Mendoza. deArgentina. garbanzo 45(2): y efecto 65-73. del extracto de

In: Clobert, J.; Baguette, M.; Benton, T.; Bullock, J. (Eds) Dispersal Ecology and Evolution. 3. Benton, T. M.; Bowler, D. E. 2012. Dispersal in invertebrates: influences on individual decisions.

UK. Oxford University Press. 41-46.Pararge aegeria. Functional Ecology. 16: 484-491. 4.5. Berwaerts,Chandler, L. K.; D. Van 1985. Dyck, Flight H.; Aerts,activity P. of2002. Liriomyza Does flight trifolii morphology (Diptera: Agromyzidae) relate to flight inperformance? relationship anto experimentalplacement of yellowtest with traps the in butterfly bell pepper. Journal of Economic Entomology. 78: 825- 828.

350 Revista de la Facultad de Ciencias Agrarias Dispersal of the pea leaf miner

Oxford University Press. 462 p. 6.7. Fenoglio,Clobert, J.; M. Baguette, S.; Videla, M.; M.; Benton, Morales, T. J.G.; M. Bullock, 2017. Evaluating J.M. 2012. trap Dispersal cropping Ecology strategies and Evolution.for insect pest UK. control through simulation models. Journal of Pest Science. 90: 601-610. 8. Freeman, J. A. 1945. Studies in the distribution of insects by aerial currents-The insect populations of the air from ground level to 300 feet. Journal of Ecology. 14: 128-154.

marking frequency. Basic and Applied Ecology. 8: 377-386. 9.10. Fric, Heidinger, Z.; Konvicka, I. M. M. M.; 2007 Hein, Dispersal S.; Bonte, kernels D. 2010. of butterflies: Patch connectivity Power-law and functions sand dynamicsare invariant affect to dispersal-related morphology of the blue-winged grasshopper Oedipoda caerulescens in coastal grey dunes. Insect and Conservation Diversity. 3: 205-212. 11. Jones, V. P.; Parrella, M. P. 1986. The movement and dispersal of Liriomyza trifolii (Diptera: Agromyzidae) in a chrysanthemum greenhouse. Annals of Applied Biology. 109: 33-39. 12. Ludwig, M.; Schlinkert, H.; Meyhöfer, R. 2017. Wind-modulated landscape effects on colonization of Brussels sprouts by insect pests and their syrphid antagonists. Agricultural and Forest Entomology. doi:10.1111/afe.12237.

at Serra da Canastra National Park, Brazil. Journal of Insect Conservation. 14: 401-411. 13. Marini-Filho, O. J.; Martins, R. P. 2010. Nymphalid butterfly dispersal among forest fragments of adult pea leafminer, Liriomyza huidobrensis (Blanchard) (Diptera: Agromyzidae), in 14. Martin, A.celery. D.; Vernon, Journal R. of S.; Entomological Hallett, R. H. 2005.Society Influence of Ontario. of colour136: 25-35. and trap height on captures 15. Mazzi, D.; Dorn, S. 2012. Movement of insect pests in agricultural landscapes. Annals of Applied Biology. 160: 97-113.

Clobert, J.; Baguette, M.; Benton, T.; Bullock, J. (Eds) Dispersal Ecology and Evolution. 16. Nathan, R.; Klein, E. K.; Robledo-Arnuncio, J. J.; Revilla, E. 2012. Dispersal kernels: review. In: 17. Osborne, J. L.; Loxdale, H. D.; Woiwod, I. P. 2002. Monitoring insect dispersal: methods and UK. Oxford University Press. 187-202. British Ecological Society. 24-49. 18. Ozawa, A.;approaches. Saito, T.; Ota, In: M. Bullock, 1999. Dispersal J. M.; Kenward, of adults R. of E.; american Hails, R. serpentineS. (Eds) Dispersal leafminer, Ecology. Liriomyza UK. trifolii Entomology. 43: 49-54. 19. Parrella, M. P. 1987. (Burgess) Biology (Diptera: of Liriomyza Agromyzidae). Annual Review in fields. of Entomology. Japanese 32: Journal 201-224. of Applied

Ecosystems and Environment. 22/23: 539-554. 20.21. Pasek,R Development J. E. 1988. Core Influence Team of2015. wind R: and A Language windbreaks and on Environment local dispersal for ofStatistical insects. Computing.Agriculture R Foundation for Statistical Computing, Vienna. Available in: https://www.R-project. org/ (Accessed 15 May 2016). 22. Taylor, R. A. J. 1978. The relationship between density and distance of dispersing insects. Ecological Entomology. 3: 63-70. 23. Turchin, P. 1998. Quantitative analysis of movement. Measuring and Modelling Population Redistribution in and Plants. Sunderland, Sinauer Associates. 410 p.

oviposition drives host plant selection when parents and offspring agree on best 24. Videla, resources. M.; Valladares, Oecologia. G. R.; 169: Salvo, 743-751. A. 2012. Choosing between good and better: optimal 25. Weintraub, P.; Horowitz, A. 1995. The newest leafminer pest in Israel, Liriomyza huidobrensis. Phytoparasitica. 23: 177-184.

26. Weintraub,Johansen, P. G.; Scheffer,N.; Hallet, S. R.; J.; Civelek, Visser, D.;H.; Valladares,Chen, B.; Metzles, G.; Soares H. 2017. Correa, The A.;invasive Shepard, Liriomyza B. M.; huidobrensisRauf, A.; Murphy, (Diptera: S.; Mujica, Agromyzidae): N.; Mac Vean,understanding C.; Kroschel, its pestJ.; Kishinevsky, status and managementM.; Joshi, R.; globally. Journal of Insect Science. 17: 1-27.

27. Yoshimoto, C. M.; Gressitt, J. L. 1964. Dispersal studies on Aphididae, Agromyzidae and Cynipoidae. Pacific Insects. 6: 525-531.

Tomo 51 • N° 2 • 2019 351 M. S. Fenoglio, M. Videla, A. Salvo, J. M. Morales

28. Zehnder,Statistical G. W.; Trumble, Software, J. T. 1984. 27. Available Intercrop in:movement http://www.jstatsoft.org/v27/i08/ of leafminers. California Agriculture. (Accessed 38: 7-8. 13 29. Zeileis, JuneA.; Kleiber, 2017). C.; Jackman, S. 2008. Regression Models for Count Data in R. Journal of

30. Zuur, A. F.; Ieno, E. N.; Walker, N. J.; Saveliev, A. A.; Smith, G. M. 2009. Zero-truncated and zero- inflatedSpringer. models 261-293. for count data. In: Zuur, A. F Ieno, E. N.; Walker, N.J.; Saveliev, A. A.; Smith, G. M. (Eds) Mixed effects models and extensions in ecology with R. New York. Acknowledgements

Ciencia, Tecnología e Innovación Productiva-Argentina) for the grant for this research (PICT We thank to FONCyT- Agencia Nacional de Promoción Científica y Tecnológica (Ministerio de was performed. We would like to thank to two anonymous reviewers for helpful comments 2012-No.on the manuscript. 1846). Special M. S. thanksF., M. V., to A. Gerardo S. and J. Videla, M. M. belong owner to of CONICETthe field where (Consejo the Nacional experiment de

Investigaciones Científicas y Técnicas).

352 Revista de la Facultad de Ciencias Agrarias