The best approach would be to remove the unconsumed feed every time you administer a new portion. In that way you can be (almost) sure, that decomposition/alteration of feed does not influence  your measurement of the feed consumption. Every other schedule might result in questions of how leftover feed might have influenced the results.

Dear Gielo,

it depends on your question and how much time you have. In general, everything works as far as you describe it in the "methods" section of your paper.

I put an example in the link below. This was a feeding experiment, where assimilation of 11 chemical elements was measured, including trace elements. Since the concentrations of some of the elements could be influenced by microbial action, in this case the author decided to collect unconsumed feeds and feces every second day. But even once a week may be enough if only results wont be biased because of this.

Kind regards!

Michał Filipiak 

Data M Asselman Poster Feeding setup Puławy

Thanks @Mr. Kan and Filipiak. 

Whether using restricted  or ad libitum feeding, frequency of feeding may determine voluntary dry matter intake, which you will then use as a basis for judging whether a feed is more acceptable than the other or not. The voluntary dry matter intake should then be evaluated  on a 24 hour cycle basis. Preferably collecting the refusals once every 24 hours will provide better results whose means over several days you will compute. But the frequency of providing the feed to the animal  within the 24 hour cycle will also greatly influence voluntary DM intake and level of selection. In addition the order in which you provide the feed ingredients  if they are not a total mixed ration (TMR) as well as whether there is some completion in eating created by one animal seeing the other eating will greatly matter.

A 24 hour cycle and a strict daily shedule is my standard recommendation.

Dear Gielo Manguiran,

my experience is restricted to ruminants.

The protocol used by our research group is as follows:

First of all we provide a feed quantity 5% higher than the expected consumption to avoid selection by animals, (the expected dry matter consumption can be assessed in publications from NRC) ; 2nd Collect all uncosumed feed (leftover) in 24 h; 3rd analize chemical composition from the leftover and compare it to provided feed.

Making the steps above we can measure the comsumption and the ingredient selection.

Thanks for your answers. :)

I'm confused on what to do with the unconsumed feeds from the first feeding. Should I collect twice a day then?

@Mr. Correa, unfortunately, I wouldn't be able to do much analysis as it is wickedly costly here. I was advised to just go with my data and use the formula:

Acceptability(%)=𝐹eed given- Feed consumed

                                / Feed given and multiplied by 100

Step by step you will build an own useful and comparable protocol. Both animal digestive system and symbiotic population need time in order to manage any change in diet composition. Your protocol must consider that fact. Also you must be aware of the protocols used in those trials you want to use as comparison or references.

Research is a high cost labor and results value higly depend on its repeatability (here is where protocol is paramount).

Mr. Correa example is a good, simple and easy to use protocol.