1 Department of Botany, University of Manitoba, Winnipeg, Manitoba R3T 2N2, Canada; 2 Department of , University of Guelph, Guelph, Ontario, Canada

Background and objectives The dwarf mistletoes are the second most important cause of timber loss in western North America. Arceuthobium americanum, causes witches' brooms, deformation and reduction of vigour in jack pine in the Canadian prairie provinces . Its only means of spread is by seeds which are explosively discharged up to 18 m. [1] and carried by birds over longer distances. Available inoculum is determined by the ability of the fruit to develop to maturity. This study was designed to assess the factors that contribute to fruit mortality during the 15 months between fertilization and seed discharge, and to test the hypothesis that fruit mortality is correlated with moisture stress.

Materials and methods
Observations were made on a random stratified sample of 300 pistillate dwarf mistletoe shoots over 2 seasons in a heavily infected jack pine stand. Fruit and shoot mortality, occurrence of resin disease and Walrothiella arceuthobii, temperature and vapour pressure deficit (VPD) were recorded.

Results and conclusions
Live fruit retention was consistently high during both winters, dropping slightly in October but remaining above 85% until spring; by June it had droppped to around 75%. In the first season it reached 60% in July and 30% just prior to seed discharge in August. The following season it had dropped to 30% by July and no fruits remained at the August sampling time after seed discharge had begun. Survivorship curves thus conformed to Type 1 of Pearl [2]. When instantaneous mortality rates were plotted against VPD and Pearson product moments calculated, a positive correlation emerged, insignificant in 1993 (r=0.38) but highly significant in 1994 (r=0.97). In 1994, when VPD frequently exceeded 1 KPa, moisture stress appeared to be the key factor in mortality, whereas in 1993, when it rarely did so, other factors were more important. W. arceuthobii accounted for a maximum of 20% and resin disease 10% (both in May 1993) of fruit mortality in any period but neither was significantly correlated with VPD. Individual fruit mortality reached 46% in June 1993 and 61% in July 1994. In 1993 it accounted for more than 20% of the loss in each of the summer months and was significantly correlated (r=0.91) with VPD. However in 1994 the correlation was not significant as it was only a major factor in June, causing 61% of mortality in that month. Death and secession of entire aerial shoots was the main cause of fruit failing to mature in both seasons. In 1993 it accounted for >40% of fruit loss in each of the summer months and in 1994 for %gt;80% in July and August; significant correlations with vpd were r=0.87 and r=0.92 respectively. One may conclude that the loss of approximately 90% of the fruits between fertilization and maturity is largely due to shoot mortality brought about by high evaporative demand when VPD is consistently above 1 KPa. In months and seasons with less moisture stress, other factors such as fungal parasitism, and perhaps insect attack, may also contribute to individual fruit mortality. Prediction of overall inoculum levels is further complicated by the uneven distribution of shoots and developing fruits within the host population. This is probably a reflection of the differing physiological vigour of individual trees and hence their ability to supply the transpirational demands of the parasite. The result is an aggregated distribution of the mistletoe which is characteristic of a steady state relationship between host and parasite and may help to explain the low rate of increase in mistletoe populations.

1. Punter D, Gilbert J, 1991. Canadian Journal of Forest Research 21, 434-38.
2. Pearl R, 1928. The rate of living. Knopf, New York.