High Tunnels and Secondary Row Covers Both Significantly Alter Winter-Time Air Temperatures and Growing Degree Day Accumulation in a Southern Trial

High tunnels are unheated greenhouse-like structures used for season extension, in which temperature is often controlled by opening and closing walls and doors.  A study is underway at Crystal Springs, Mississippi testing the influence of 5.0 ft tall, 3 ft. wide over-the-row tunnels (secondary covers) within high tunnels on the tunnel microclimate and crop performance.  We established tomato (Lycopersicon lycopersicum cv. Legend) and zinnia (Zinnia elegans cv. Benary’s Giant Crimson) transplants into replicated plots in Fall 2009 in three high tunnels.  Cover treatments within the tunnels included no cover (NNCOV); and four covering fabric treatments that could be pulled over or off of a frame built over the crop row.  A fixed protocol based on actual and forecast outside and tunnel temperatures guided tunnel opening and closing, as well as placement and removal of the secondary covers.  We recorded air temperatures inside (NCOV) and outside of the tunnels (OUT), and under each cover treatment (COVS) in triplicate in each crop.  Injury from several light frost events was more evident on NCOV than COVS plots, but no treatment prevented all injury from severe events in early January.  For this first year of study, we could not complete maturation of a tomato crop before three days in a row with temperatures falling below 18°F ruined it.  Up to that point, at which we terminated the experiment, the presence of the tunnel influenced air temperature more than covers.  Compared to OUT, NCOV lead to an average increase in minimum, maximum and average temperature of 6.2, 29.3, and 9.3°F, respectively.  Compared to NCOV, COVS lead to an increase in minimum, maximum and average temperature of +1.0 to +2.4, -0.3 to +0.2, and +0.9 to +2.4°F, respectively, within the tunnels.  Growing degree-day (GDD) accumulation inside the tunnels, at GDD, base 40°F was more than 200% by greater than OUT, more than 500% at GDD₅₀.  The GDD₄₀ NCOV averaged 5% lower, and the GDD₅₀ 15% lower than with COVS.