Coccidiosis in chickens: From temporal patterns of Lesion development and Oocyst shedding to predictive modeling.

Coccidiosis continues to be an impactful parasitic disease in the poultry industry, and the increasing resistance to conventional anticoccidials such as ionophores and chemicals underscores the necessity of a deeper understanding in the behavior of this parasite. These experiments aimed to evaluate the distribution of lesion scores, oocysts shedding, growth performance, and to develop coccidiosis prediction models in broiler chickens undergoing a coccidiosis challenge and supplemented or not with ionophores or chemical/ionophore blend. Three experiments were conducted with 1,200 (Exp. 1) and 800 (Exp. 2 and 3) Cobb 500 male chickens in a completely randomized block design. Treatments included: (1) Control, (2) Challenged, (3) Challenged supplemented with ionophore (Salinomycin, SAL) at 66 ppm, and (4) Challenged supplemented with a blend of chemical/ionophore (Maxiban®, MAX) at 72 ppm. Birds were raised on new litter with 12 (Exp. 1) or 8 (Exp. 2 and 3) replicate pens/treatment. Coccidia challenge involved E. acervulina (150,000), E. maxima (50,000), and E. tenella (150,000) oocysts (Exp. 1 and 2), and E. acervulina (150,000), E. maxima (75,000), and E. tenella (150,000) oocysts (exp. 3), given via oral gavage at d14. The studies lasted 35d. Birds were necropsied at d19, 20, and 21 for lesion score evaluation, and excreta samples collected for oocyst per gram (OPG) analysis during two cycles of the coccidia (d19-22 and d26-28). Growth performance (body weight gain, BWG; feed intake, FI; and feed conversion ratio, FCR) was assessed weekly and from 0 to 35d. Overall, from d1-35, both SAL and MAX improved FI, BWG, and FCR (P < 0.01) while partially reducing the mortality rate (P < 0.05). Regardless of anticoccidial treatment, the peak of lesions of E. acervulina was on d19, and E. maxima and E. tenella on d20 with a reduction in severity of lesions afterwards. The peak of oocysts shedding was at d20 (1st cycle) and at d27 (2nd cycle). Both treatments (SAL and MAX) attenuated the coccidiosis lesions of the three Eimeria species and OPG (P < 0.01) during the three timepoints evaluated. The prediction models identified total OPG as key predictor for lesion severity. BWG at d28 was associated with E. acervulina and E. tenella lesions, and total OPG at d20 and d28, and BWG at d35 with E. acervulina and E. maxima lesions, and total OPG at d19, d20, and d28. In conclusion, this study demonstrates the value of frequent, time-sensitive measurements and the potential of data-driven approaches. Hence, the integration of classical statistics with machine learning techniques may be essential for improving the understanding of coccidiosis in broilers.