Silva, J. L.Guim, A.Carvalho, F. F. R.Mattos, C. W.Menezes, D. R.Coelho, M. C. S. C.Garcia, D. A. [UNESP]Pereira, I. G.Pereira Neto, J. D.2018-11-262018-11-262015-12-01Small Ruminant Research. Amsterdam: Elsevier Science Bv, v. 133, p. 10-16, 2015.0921-4488http://hdl.handle.net/11449/158635Factor analysis was used for reduction of dimensionality and to obtain latent variables, constructed from linear combinations of feeding behavior characteristics, physiological and performance, to evaluate the effects of diets containing integral mango meal (IMM) replacing corn (0, 330, 660, and 1000 g/kg) for dairy goats. Eight lactating Saanen crossbred goats (48.7 +/- 1.99 kg of BW) were introduced in the experiment 48 days postpartum and maintained up to 124 days of lactation. The latent variable composed of behavioral characteristics and milk performance showed a cubic effect for the tested diets (P<0.05), with a maximum point for milk performance with 78 g/kg of replacing corn with integral mango meal, and a minimum point with 763 g/kg replacing corn with integral mango meal. Similarly, there was a linear increase (P<0.01) in rumination and chewing activities with replacement levels. The latent variable composed of metabolic characteristics and milk performance did not (P > 0.05) detect the effect of the replacement of corn with integral mango meal. From the factor analysis, it was possible to reduce the dimensionality of the set of data. Considering that the feeding behavior characterized by the amount of meals and activities of rumination interfered the performance of dairy goats, may be recommended replacing 78 g/kg of corn by integral mango meal. (C) 2015 Elsevier B.V. All rights reserved.10-16engAlternative feedFactor analysisIngestive behaviorMetabolic profileMilk productionArtigo10.1016/j.smallrumres.2015.10.023WOS:000367109700003Acesso abertoWOS000367109700003.pdf