Dry matter intake (DMI)

Definition

Dry matter intake (DMI) is the weight of feed dry matter a cow eats per day, with the water removed. It is stated in lb DM per cow per day (or kg), and for a group it is the pen total divided by the average number of cows in the pen.

Two numbers get called DMI, and they are not the same:

DM delivered (or “DM fed”): the as-fed weight of TMR dropped in the bunk times the TMR dry matter fraction. This is what feeding software reports by default.
DM consumed: DM delivered minus the DM in refusals (feed weighed back out of the bunk, usually before the next morning’s feeding). This is the true intake and the one to use for feed efficiency.

The gap between the two is refusals. Refusals that are thrown out, rather than re-fed to heifers or other groups, are part of shrink.

DMI is tracked by pen, because pens get different rations and cows in them differ in stage of lactation and parity. A herd average hides a fresh pen that is under-eating.

Measuring DM% of the TMR

DMI is only as good as the dry matter value used to convert as-fed pounds.

Oven or microwave drying of a sample on farm (microwave or Koster tester, forced-air oven), or a lab analysis. On-farm tests are quick and good enough for weekly ration adjustments.
Sample forages, not just TMR. Wet forages (silages, wet byproducts) drive TMR DM swings. Herds commonly test silage DM at least weekly and after rain, and adjust the as-fed recipe to hold DM delivered constant.
Calculated TMR DM (from ingredient DM values in the ration software) is an estimate. A measured TMR sample checks it.

Formula

Plain text:

DM delivered (lb/cow/day) = TMR as-fed delivered (lb) × TMR DM% / 100 / cows in pen
DM refused   (lb/cow/day) = refusals as-fed (lb) × refusal DM% / 100 / cows in pen
DMI consumed (lb/cow/day) = DM delivered − DM refused
Refusal rate (%)          = refusals as-fed (lb) / TMR as-fed delivered (lb) × 100

LaTeX:

\mathrm{DMI} = \frac{W_{\text{fed}} \cdot d_{\text{TMR}} - W_{\text{ref}} \cdot d_{\text{ref}}}{N_{\text{cows}}}

where W = as-fed weight in lb, d = dry matter fraction (DM% / 100), and N = average cows in the pen for the day.

Refusals are usually wetter or drier than the TMR as fed (cows sort out long, dry particles or leave fines). Use a refusal DM% if one is measured; using the TMR DM% is a common approximation.

Expected DMI

NRC (2001) predicts DMI for lactating Holsteins from milk and body weight:

DMI (kg/day) = (0.372 × FCM + 0.0968 × BW^0.75) × (1 − e^(−0.192 × (WOL + 3.67)))

\mathrm{DMI}_{\text{kg}} = \left(0.372 \cdot \mathrm{FCM} + 0.0968 \cdot \mathrm{BW}^{0.75}\right)\left(1 - e^{-0.192\,(\mathrm{WOL} + 3.67)}\right)

where FCM = 4% fat-corrected milk (kg/day), BW = body weight (kg) and WOL = week of lactation. This is Equation 1-2 in NRC 2001, Chapter 1. The last term lowers predicted intake in early lactation. The equation has no temperature or humidity adjustment. NASEM (2021) replaced it with equations that also use parity and diet fiber.

A quick rule of thumb: mid-lactation Holsteins eat roughly 3.5–4.0% of body weight as DM per day.

How to pull it in DC305

Feed weights are not in DC305. Delivered and refused weights come from feeding software (TMR management systems that log mixer-scale weights by pen) or from a manual feed sheet. TMR DM% comes from on-farm moisture tests or the lab.

DC305 supplies the denominator: how many cows were in each pen.

COUNT BY PEN FOR LACT>0          cows per pen today
SUM BY PEN FOR LACT>0            head count by pen (SUM with no items counts animals)
SUM MILK DIM BY PEN FOR LACT>0   milk and stage of lactation by pen, for expected DMI

DC305’s count is a point-in-time snapshot. Cows move between pens during the day; for daily DMI, use the head count the feeding software was given (often a manual pen count entered at feeding), or an average of pen counts across the period. Make sure the pen numbers in the feeding software match DC305 PEN values.

Typical range

Herd context: US Holsteins, freestall or dry-lot, lactating cows fed a TMR, DM consumed (after refusals).

Group	Typical DMI (lb DM/cow/day)
Lactating herd average	50–58
High-producing pen, mid-lactation	58–65+
Fresh pen (0–21 DIM)	38–48, rising each week
Close-up dry cows	25–30
Far-off dry cows	25–30

Refusals: lactating pens are commonly fed for about 2–5% refusals (as-fed). Zero refusals means cows ran out of feed; more than about 5–8% wastes feed unless refusals are re-fed.

Heat stress lowers DMI; intake drops of 10–25% during hot weather are commonly reported, with a lag of 1–2 days behind the heat. NRC 2001 cites DMI falling above about 20°C, by 22% in multiparous and 9% in primiparous mid-to-late-lactation cows under heat stress (Holter et al. 1997), and by up to 55% in severe cases.

Jerseys eat less in absolute terms but more as a percent of body weight.

Common mistakes

Reporting DM delivered as DMI. Without subtracting refusals, DMI and feed efficiency are overstated by the refusal rate.
Using a stale or calculated DM%. A silage that goes from 35% to 31% DM after rain cuts delivered DM by over 10% at the same as-fed weight, while the report still shows the old number.
Wrong cow count. Using the pen’s stall count, last week’s count, or DC305’s count at a different time of day than feeding. Hospital and fresh cows moving between pens are the usual source of error.
Herd average only. A fresh pen falling behind on intake is the earliest sign of transition trouble, and it disappears in the herd number.
Mixing as-fed and DM. Comparing an as-fed intake (often 100–130 lb) to a DM benchmark, or the reverse. See as-fed vs dry matter.
Ignoring mixer-scale accuracy. DMI inherits every error in the loaded and delivered weights. See load accuracy.