Eucalyptus is the most widely planted exotic species in the tropics, and play important roles in reforestation and production of timber, pulp, potential bioenergy feedstock, and other forest products ( Wyle & Floyd, 1998 ; Rejmánek & Richardson, 2011 ). In Argentina, Eucalyptus spp. plantations occupy about 261,000 ha, mainly in the northeast region ( MAGyP, 2015 ).

The eucalyptus gall wasp, Leptocybe invasa Fisher & La Salle (Hymenoptera: Eulophidae), native from Australia, is a worldwide pest in Eucalyptus plantations. L. invasa has expanded to more than 29 countries in Asia, Europe, Africa and the Americas ( Mendel et al. , 2004 ; Zheng et al. , 2014 ). This insect was first repor-ted in Argentina in 2009 ( Aquino et al. , 2011 ). The fe-male L. invasa lays her eggs in plant tissues causing the formation of galls on the leaf midribs and petioles and on the stem of new shoots, eventually leading to leaf-curling and premature aging of the leaves. Eggs overloading might cause death of youngs shoots, while severe attacks lead to leaf fall, stunted growth and may seriously weaken the tree ( Mendel et al. , 2004 ). Gall development occurs as response to a physiological di-sorder caused by a specific relationship between plant and insect. It is highly dependent on specific small cha-nges in morphological and phenological characteristics of the plant, which determines the degree of suscepti-bility and coevolution between plant and insect ( Stone & Schönrogge, 2003 ; Zheng et al. , 2014 ).

Eucalyptus species show different susceptibility to L. invasa attacks ( Mendel et al. , 2004 ) with Eucalyptus grandis W. Hill, Eucalyptus camaldulensis Dehnh and Eucalyptus tereticornis Smith being the most susceptible ones ( Thu et al. , 2009 ). However, the susceptibility can be strongly influenced by local environmental factors, and therefore further susceptibility tests are needed in locations where eucalyptus trees are cultivated. The ob-jective of the present study was to evaluate the suscep-tibility of different Eucalyptus species and hybrids to L. invasa planted in the field in Argentina.

This study was conducted from March to November 2014 in Colonia Delicia, Misiones, Argentina (26°10'16" S, 54°33'15"W). The clime is subtropical without dry se-asons. Eucalyptus seedlings, less than one year old, were planted in a field experiment without irrigation. L. invasa damages were evaluated in a 1.35 ha field plot, with a spa-cing between plants of 3.0 × 2.5 m (1,333 plants/ha). The evaluated species were E. camaldulensis, E. grandis, E. major, E. propinqua, E. tereticornis, E. dunni, E. molucca-na, E. longirostrata, E. urophylla and the hybrids E. gran-dis × E. camaldulensis and E. urophylla × E. grandis (Table 1), planted in October 2013.

The experiment was designed as a complete rando-mized block, with 4 blocks and 5 replications. In each block, 11 experimental plots of each species with 5 plants were randomly stablished, and all trees in these plots were surveyed for damage and severity of L. invasa infestation. Six evaluations were made during 2014 between March and November, evaluating the damage index (DI) and average damage index (ADI). DI scored the proportion of leaves and twigs damaged per tree, as follows: 0, no damage; 1 < 25%; 2, 25-50%; 3, 51-75%; 4 >75% ( Thu et al. , 2009 ). DI was calculated as the percentage of trees affected. ADI was calculated according to:

ADI= Σ ni·vi / Ni

where ni is the number of trees infected at DI= i , v _i the da-mage index at level i , and Ni is the number of trees assess-ed per species. Based on maximum ADI values, a damage severity level was defined ( Thu et al. , 2009 ): nil (ADI = 0), low (ADI: 0-1.0), medium (ADI: 1.1-2.0), severe (ADI: 2.1-3.0) and very severe damage (ADI: 3.1-4.0).

Statistical analysis was performed using analysis of variance (ANOVA) after checking that normality and homoscedasticity were fulfilled, and means were compared with the Tukey test ( p <0.05), using InfoStat statistical software V 2016 ( Di Rienzo et al. , 2016 ).

Damage incidence by L. invasa significantly differed among the tree species ( df =18, F =28.42, p <0.0001). The highest damage incidence was observed on E. tereticornis and the lowest in E. urophylla, E. urophylla × E. grandis and E. moluccana. E. grandis × E. camaldulensis, E. lon-gistrata, E. major, E. grandis, E. camaldulensis, E. dunnii and E. propinqua showed medium damage incidence (Table 2).

Based on ADI, in the six evaluations between March and November 2014, the resistance of eucalyptus species differed widely. E. urophylla and the hybrid E. urophylla × E. grandis seemed tolerant to L. invasa and no galls were produced. Six species ( E. moluccana, E. longistrata, E. major, E. grandis, E. camaldulensis, E. dunnii ) and the hybrid E. grandis × E. camaldulensis showed low damage severity. Medium damage was ob-served in one species, E. propinqua and severe damage occurred in E. tereticornis (Table 2).

Differences in susceptibility to the gall-forming insects of eucalyptus tested may indicate that genetic factors are involved in both, attractiveness to L. invasa for oviposition and suitability for L. invasa larval development. Similarly, synchronization of gall-forming species with host plant phenology is behind many cases of susceptibility as consequence of coevolution between eucalyptus plants and L. invasa . These factors open the chance of coping with this pest through the development of resistant and productive genetic stock, better than through chemical means ( Dungey et al. , 2000 ; Guerreiro et al. , 2015 ). In this context, the order of susceptibility to L. invasa in Misiones was E. tereticornis > E. propinqua > E. dunni > E. camaldulensis > E. grandis > E. major > E. longistrata >E. grandis × E. camaldulensis . Other studies have also showed susceptibility to L. invasa of these species ( Mendel et al. , 2004 ; Thu et al. , 2009 ; Javaregowda & Prabhu, 2010 ; Zhu et al. , 2012 ). On the other hand, E. moluccana, E. urophylla × E. grandis and E. urophylla resulted tolerant to L. invasa , as previously shown by Thu et al. (2009) , Nyeko et al. (2009) and Guerreiro et al. (2015) .

This study suggests that in Misiones, E. tereticornis is the most sensible eucalyptus species to gall wasp attack, whereas other species and one hybrid of regional economic importance like E. dunni, E. camaldulensis, E. grandis and E. grandis × E. camaldulensis presented low damage levels. The hybrid E. urophylla × E. gran-dis showed tolerance to L. invasa . More studies are ne-cessary to determine the tolerance to L. invasa of dif-ferent eucalyptus species and hybrids cultivated in the northern Argentine region.