Using a Segmented Model to Describe In situ Nutrient Disappearance

Stacey A. Gunter, Michael L. Galyean

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


The purpose of this study was to compare the predictive results and characteristics of exponential (EM) and segmented models (SM) describing ruminal in situ nutrient disappearance data. Using masticate samples collected from esophageally cannulated steers grazing midgrass prairie rangeland (PRAIRIE) or a sideoats grama (Bouteloua curtipendula [Michx.] Torr.)/sweetclover (Melilotus officinalis [L.] Lam.) pasture (PASTURE), in situ organic matter (OMD) and nitrogen (ND) disappearance curves were fitted to a single EM (Rt=D0*e(-k[t-L])+I0) on a SM. The SM contained a quadratic segment (Y=β01X+β2X2), where digestion seemed to occur and after the point where digestion had seemed to cease a constant (Y=plateau) was fitted. With the SM, linear (β1) and quadratic (β2) %/h) effects of OMD and ND did not differ (P>0.15) between PRAIRIE and PASTURE. With the EM, the rate of OMD (%/h) was 44% faster (P=0.04) for PRAIRIE than for PASTURE, but disappearance rate did not differ (P=0.24) for ND. The hour where the plateau joined the quadratic equation in the SM did not differ (P=0.31) between forage types (42 and 41 h for OM and N, respectively). Soluble OM and N fractions did not differ (P>0.99) between models; however the soluble OM and N concentrations were greater (P<0.004) for PRAIRIE than for PASTURE. Estimates of insoluble degradable OM and N fractions did not differ (P>0.42) between models. The insoluble degradable OM in PRAIRIE did not differ (P=0.52) from PASTURE; whereas, the insoluble degradable N in PASTURE was greater (P=0.08) than in PRIARIE. Indigestible OM and N fractions did not differ (P>0.11) between models; but the indigestible OM and N was greater (P<0.004) for PASTURE than for PRAIRIE. Using either the SM or EM revealed that OM and N in PRAIRIE masticate tended to degrade at a faster rate and was more digestible than OM and N in PASTURE. Based the correlation analysis, the SM model was generally more accurate and displayed less bias than the EM for both in situ OMD and ND, but differences were extremely small. Therefore, either model may be used when attempting to develop empirical models to predict the extent of in situ digestion.

Original languageEnglish
Pages (from-to)1-14
Number of pages14
JournalJournal of Applied Animal Research
Issue number1
StatePublished - Sep 2000


  • Non-linear regression
  • Ruminal fermentation
  • Ruminally degraded nitrogen
  • Segmented models


Dive into the research topics of 'Using a Segmented Model to Describe In situ Nutrient Disappearance'. Together they form a unique fingerprint.

Cite this