A data base was constructed to describe and estimate supplementation effect
s in nonlactating cattle consuming forage ad libitum. The data base include
d 66 publications on 126 forages (73 harvested and 53 grazed) and a total o
f 444 comparisons between a control, unsupplemented treatment and a supplem
ented treatment. Daily gains were reported for 301 comparisons and voluntar
y intake for 258. Direct measures of forage digestibility were reported for
202 comparisons, and total diet digestibility for 150. Supplements did not
increase gain in all cases. Change in ADG due to supplement was not relate
d closely to intake of supplemental TDN. Lowest increases in ADG were with
native forages supplemented with molasses alone or with low intakes of mola
sses containing high levels of NPN. Greatest increases in gain were with im
proved forages, supplements with > 60% TDN, and supplemental CP intake > .0
5% of BW. Supplements decreased voluntary forage intake (VFI) when suppleme
ntal TDN intake was > .7% of BW, forage TDN:CP ratio was < 7 (adequate N),
or VFI when fed alone was > 1.75% of BW. When supplements increased VFI, fo
rage TDN: CP ratio was > 7 (N deficit), and VFI when fed alone was often lo
w. There was little relationship between change in VFI and sources of suppl
emental CP and TDN. Supplements caused total diet TDN concentration to devi
ate from expected values by -10 to +5% of OM. When supplemental TDN intake
was > .7% of BW, diet TDN concentration was always less than expected. Ther
e was little relationship between deviation from expected total diet TDN an
d type or composition of forages or supplements. Empirical multiple regress
ion equations were developed to estimate effects of supplements on VFI and
total diet TDN concentration. The most acceptable intake equation estimated
VFI when fed with supplement (r(2) = .84) That equation included VFI when
fed alone, supplement intake, CP and TDN concentrations in forage and suppl
ement, and classification codes describing forages and supplemental energy.
The most acceptable equation for estimating total diet TDN concentration i
ncluded only the expected total diet TDN concentration (r(2) = .87). These
equations may be used in nutritional models to account for associative effe
cts.