Host-preference and density of woodrose-forming mistletoes (Loranthaceae) on savanna vegetation,South Africa

In the Bushbuckridge region of South Africa host preference and density of two woodrose-forming mistletoes, Erianthemum dregei (Eckl. & Zeyh.) V. Tieghem and Pedistylis galpinii (Schinz ex Sprague) was quantified in relation to land-use (harvested or unharvested), rainfall (high > 660 or low < 660 mm year^–1) and catenal position (top or lower slope). These two mistletoes are generalist hemi-parasites of savanna trees and shrubs occurring on 25 and 17 hosts respectively, seven of which are shared. Thirty-six percent of woody plant species recorded were found to be hosts. Although Sclerocarya birrea (A. Rich.) Hochst. comprised only 4% of woody plant density in the environment, it was the principal host for both mistletoes, accounting for 71% of total E. dregei and 42% of P. galpinii infection. Mistletoe infection relative to density of Ficus stuhlmanii, Trichilia emetica and Cassine transvaalensis indicated that these were preferential hosts to S. birrea. Mistletoe host preference was negatively correlated with host wood density. Mistletoe number per tree had a weak relationship to canopy size. Mistletoes of all size classes were denser at high rainfall relative to low rainfall sites. Interestingly, the overall mistletoe size class distribution was similar between harvested and unharvested sites. The ratio of living to dead mistletoe was 2 to 1 for E. dregei and 1.5 to 1 for P. galpinii. There are sufficient dead mistletoes in unharvested and harvested areas to satisfy present market demand. Living E. dregei predominated in harvested rather than unharvested areas suggesting that current-harvesting levels had little or no negative effect on the population. In contrast, P. galpinii was denser in unharvested areas possibly owing to its higher market value and thus higher harvesting levels.