Get started
Work

PSX-20 Cluster analysis to identify bone quality patterns in broilers fed with Trisetum flavescens extract.

Year: 2025
Type: article
Abstract: This study aimed to use cluster analysis to identify patterns in broiler bone variables related to the addition of Trisetum Flavescens extract. Bone quality significantly affects broiler performance and welfare. Various nutritional strategies can affect bone weight, length, strength, ashes, and dry matter. While ANOVA is commonly used for comparing treatments, other methods like K-means clustering can reveal hidden patterns in data by grouping similar individuals. A total of 72 batches, in field, of broilers were used in this trial, 35 batches that received 1,25-dihidroxicolecalciferol (Active D), and 37 batches that received conventional additive program (CAP). Each batch were obtained 10 right tibiae of ten broilers chose randomly a total of 370 tibiae were analyzed by conventional additive program and 350 for 1,25-dihidroxicolecalciferol (Active D). The tibias were tested for breakage resistance using Brookfield’s CT3 Texture Analyzer, weighed on an analytical balance, and analyzed for dry matter and ash content. The dry matter content was calculated as the ratio between the weight of the tibia after oven and the weight before drying. The tibiae were weighed and measured length, after the bones were burned in muffle 600 ºC for 8 hours to obtained ashes. The variables bone length (mm), weight of bone (g), resistance to breakage (kgf), dry mater (%), and ashes in dry matter (%) were grouped by K-means method, the optimal number of clusters was determined using the Elbow Method. Following the clustering process, three clusters were identified, accounting for 45% of the total data variability. The clustering results were visualized using the first two principal components, which explained 75% of the variance. The means for clusters 1, 2, and 3 were as follows: bone weight (13.03, 13.12, 18.18), bone length (99.15, 96.16, 105), resistance to breakage (31.07, 35.12, 45.46), dry matter (47.9, 44, 45.8), and ashes in dry matter (43.3, 47.4, 45.7), with 209, 259 and 249 bones grouped by each cluster. Cluster 1 consists of bones that are weaker and have lower ash content. Cluster 2 includes smaller bones with higher ash content. Cluster 3 includes bones that are heavier, stronger, and longer. Following the results, this cluster was associated with treatments Active D and CAP, as determined by the chi-square test (χ2=21.395, p< 0.001). Specifically, Cluster 1 consisted of 38.8% bones from Active D and 61.2% from the CAP, Cluster 2 had 52.9% bones from Active D and 47.1% from the CAP, while Cluster 3 comprised 60.2% bones from Active D and 39.8% from CAP. The results of Cluster analysis indicate that Active D led to heavier, stronger, and longer bones, demonstrating its potential impact on broiler bone development.
DOI: 10.1093/jas/skaf300.668
Language: English
Open Access status: bronze
OpenAlex ID: W4414831000
FWCI:
Cites:
Cited by:
Related to: