If the total protein concentration per lane increases (confirmed by Ponceau) why would the ECL signal for the protein of interest decrease? No need to do density analysis, quite visible by direct observation.
It's not clear how you performed your western blot. Did you intentionally load some lanes with a higher total protein concentration? Are all of your lanes from the same experimental group?
I would imagine membrane binding of your protein of interest could be competing with other abundant protein with similar mobility. If you have more total protein, you will have more competitors. Since the speed of migration of proteins out of gel is not uniform, the local area of the membrane could be saturated by abundant, fast-migrating proteins.
Another potential reason is that, your protein of interest it too abundant and rapid ECL reaction (it generates reactive oxygen) has killed HRP, which is a well known phenomenon. In this case, you may observe "hollow" bands (=middle of the band has weaker signal then the outer edge).s
overloading of gel often results in 'ghost bands' which might looks fainter. Make sure you are not visualizing ghost bands. If you are observing multiple bands with increase in protein amount, the non-specific binding could be a possible reason for weakening of specific signal.
Hi Khalid: yes, I wanted to see my range and limit of detection from the same sample.
Hi Shin: I thought about your first suggestion, was not too sure, thank you for confirming my suspicion. I don't think your second suggestion is the case here, as I had to use pico-ECL and long exposure to detect it.