Testing the Proposed Connection between Dark Energy and Black Holes

In 1997, an extension of general relativity was proposed that predicts the dark energy density \Lambda\ to vary linearly with the total number of macroscopic black holes in the universe. We explore this prediction and find that \Lambda\ must be roughly constant after the bulk of the stellar mass black holes are in place, so for a redshift z smaller than unity. Conversely, the highest black hole formation rate corresponds to the peak in the cosmic star formation history, earlier than z=1. This yields a fast declining \Lambda, by a factor of about 5, from redshift 1 to 3. At even earlier times, before many stars were formed, the value of \Lambda\ should be much smaller than its current value. These predicted effects are all consistent with current data, and near future observations can definitively confirm or disproof the link between the dark energy density and the total number of black holes in the universe.