Abstract The aims of this work were to characterize the damage caused by Paraulaca dives in feijoa (Acca sellowiana) and develop a diagrammatic scale to evaluate the severity of herbivory. The evaluations were carried out in a feijoa progenies orchard with eight-year-old. The beetles caused damage mainly on leaves, from October to December (spring season). The distribution of beetles in the orchard occurred randomly, not following a defined pattern of incidence. The diagrammatic scale of herbivory severity was represented by seven levels (1%, 3%, 5%, 7%, 15%, 32%, and 55% of the leaf area consumed). The use of this diagrammatic scale significantly increased the precision and accuracy of severity estimates by inexperienced evaluators. This can help in strategies to control this pest to expand the cultivation of feijoa in Brazil. Acca sellowiana eightyearold. eightyearold eight year old. old eight-year-old leaves spring season. season . season) randomly incidence 1%, 1 (1% 3 3% 5 5% 7 7% 15 15% 32 32% 55 consumed. consumed consumed) evaluators Brazil 1% (1 (

RESUMO A área foliar (cm2 por folha) e o conteúdo de pigmentos foliares são importantes características usadas para melhor entender a fisiologia das plantas. Neste estudo, modelos não-destrutivos para área foliar e conteúdo de pigmentos baseados no comprimento (L) e na largura (W) da folha, e na leitura do teor de clorofila, foram desenvolvidos para feijoa (Acca sellowiana). O experimento foi desenvolvido durante o mês de janeiro de 2016 usando 60 genótipos com cinco anos de idade, cultivados no estado do Paraná, Brasil. A proposta de modelo para área foliar (LA) foi L A = 0 , 0022 L 3 + 0 , 1482 W 2 + 0 , 6159 L W + 0 , 1076 (R2 = 0,99). Três modelos encontrados na literatura também foram analisados. Todos esses modelos tiveram menor acurácia que o modelo proposto por este estudo. Os modelos propostos para estimar os pigmentos foliares foram baseados no Falker Chlorophyll Index para clorofila a (A) and b (B), sendo estes C h l a = 2 , 564 A + 13 , 098 B - 42 , 605 (R2 = 0,94), C h l b = 1 , 538 A + 3 , 287 B + 8 , 847 (R2 = 0,86) e C a r o t e n o i d e s = 0 , 947 B + 8 , 943 (R2 = 0,88) expressos como µmol m-2 de lâmina folha. Conclui-se que os modelos propostos neste estudo se mostraram confiáveis métodos não destrutíveis para estimar a área foliar e o conteúdo de pigmento foliares em A. sellowiana.

ABSTRACT Leaf area (cm2 per leaf) and leaf pigment content are important traits that can be used to better understand a plants physiology. In this study, empirical non-destructive models for leaf area and leaf pigment based on the leaf dimensions, length (L) and width (W) in centimeters, and chlorophyll meter readings were developed for feijoa (Acca sellowiana). The experiment was carried out during January 2016 using five-year-old trees of 60 genotypes, grown under field conditions in the state of Paraná, Brazil. The proposed leaf area (LA) model was L A = 0 . 0022 L 3 + 0 . 1482 W 2 + 0 . 6159 L W + 0 . 1076 (R2 = 0.99). Three current leaf area models found in the literature were also assessed. All of the already created models were less accurate than the model proposed in this article. The proposed leaf pigment models were based on the Falker Chlorophyll Index for Chlorophyll a (A) and b (B), these were C h l a = 2 . 564 A + 13 . 098 B - 42 . 605 (R2 = 0.94), C h l b = 1 . 538 A + 3 . 287 B + 8 . 847 (R2 = 0.86) and C a r o t e n o i d s = 0 . 947 B + 8 . 943 (R2 = 0.88) expressed as µmol m-2 of leaf blade. In conclusion, the proposed models in this study were shown to be a reliable non-destructivel way of estimating A. sellowiana leaf area and leaf pigment.