Field Crops Research, a leading agricultural journal, has just published a review of Darwinian Agriculture by Jeffrey White. He suggests that there is "ample scope for debate and further research on speciﬁc propositions of Darwinian Agriculture." I agree and look forward to those debates. For example, I argued that perennial grain crops may never approach the yields of annual grains, because perenniality requires investing resources in over-wintering that could otherwise by used for grain. White agrees that this may be true in temperate climates, but maybe not in the tropics, if the "off-season" is short. He gives "perennially cropped sugarcane fields" as an example.
I agree that perennials can potentially capture a larger fraction of annual solar radiation, as I noted in the book. For example, a time-lapse movie of rice growth shows that this "annual crop" (which may be grown 2 or 3 times in a year) completely covers the ground only a small fraction of the time. Most of the time, sunlight is mainly hitting soil, evaporating water rather than powering photosynthesis. Perennials can use some of last-season's photosynthate to power rapid leaf growth, capturing a larger fraction of solar radiation sooner.
This isn't only true in the tropics. For example, the book cites a study showing that perennial Miscanthus can produce more biomass than corn. The photo below shows the Miscanthus plots Steve Hamilton showed me at Michigan State's Kellogg Biological Station, where I gave a talk earlier this month. But this Miscanthus doesn't produce any seed and I don't think sugarcane produces much.
What about perennial grasses that do produce significant amounts of grain? Sieglinde Snapp and colleagues at Kellogg Biological Station and The Land Institute have shown that perennial intermediate wheatgrass (kernza) can reduce nitrate concentrations in soil water much more than annual wheat, presumably reducing pollution of wells and rivers. But its first-year grain yield was only 112-157 kg/ha, versus 2807-3761 kg/ha for annual wheat. (Interestingly, the higher yields were with organic management.) Second year yields were 1390-1662 kg/ha for the perennial versus 4248-5017 for the annual. Unless the perennial really improves in subsequent years, the perennial would take more than three times as much land to produce the same amount of grain. I worry that the additional land will come from clearing forests or draining wetlands.