Differential Susceptibility of 

 Ponderosa Pine to the 

 Gouty Pitch Midge 

 (Cecidomyia piniinopis) 



R. J. Hoff 



INTRODUCTION 



In 1985 a provenance test of ponderosa pine (Pinus 

 ponderosa Dougl. ex Laws.) was severely damaged by the 

 gouty pitch midge (Cecidomyia piniinopis Osten Sacken). 

 On many trees a high percentage of the current year's 

 shoots were dead or dying. Several of these trees were 

 dead a few months later. 



The gouty pitch midge (GPM) is a native pest of forests 

 of eastern and western North America. In the West the 

 pest is especially injurious to ponderosa pine where the 

 level of infestation becomes high enough to kill trees 

 (Bedard and others in press; Eaton and Yuill 1960). In 

 contrast, outside of damage caused by mortality, Bedard 

 and Ferrell (personal communication) found that growth 

 loss was not related to infestation by GPM. Natural con- 

 trols such as weather, host resistance, and parasites nor- 

 mally keep it in check (Eaton and Yuill 1960). There is 

 little doubt that the host has defense mechanisms that 

 prevent or limit infestation (Austin and others 1945; 

 DufBeld 1985; Hoff 1988). There is little information on 

 how infestation by GPM influences host resistance with 

 respect to elevation and geographic area. Ferrell and 

 others (in press) have reported that in California popula- 

 tions of ponderosa pine from the northern Sierra Nevada 

 and southern Cascades had relatively high levels of resis- 

 tance to GPM — susceptibility increased with decreasing 

 latitude. 



The purpose of this paper is to document the suscepti- 

 bility of provenances of ponderosa pine from the Northern 

 Rockies of the United States to gouty pitch midge. 



MATERIALS AND METHODS 



The study site was a provenance test that was severely 

 damaged by GPM. This plantation was established in 

 1974 by the Northern Region of the Forest Service at the 

 Lone Mountain Tree Improvement Site 25 miles (40 km) 

 north of Coeur d'Alene, ID (fig. 1). This plantation is one 

 of six that have been established in cooperation with the 

 Inland Empire Tree Improvement Cooperative. The site 

 is flat with only slight undulations at an elevation of 

 2,488 ft (758 m). The entire 160-acre (65-ha) site is sur- 

 rounded by naturally regenerated ponderosa pine and 

 lodgepole pine, with a lesser mixture of grand fir, 

 Douglas-fir, and western larch. This natural stand is two- 



layered; the overstory is composed of mature scattered 

 trees (remnants of harvest) and the lower layer is a pole- 

 sized stand 20 to 30 ft (6 to 9 m) tall. Many of these natu- 

 ral ponderosa pine were infested with GPM. 



The seed came from 92 stands (populations) located in 

 northeastern Washington, Idaho north of the Salmon 

 River, and western Montana (fig. 1). Open-pollinated 

 seed were collected from five individuals per stand. 



The seedlings were grown at a Forest Service nursery 

 near Coeur d'Alene, ID, in bare-root beds for 2 years, 

 lifted, and planted at the Lone Mountain Tree Improve- 

 ment Site in April 1974. The experimental design was a 

 randomized complete block with 10 replications (blocks). 

 Four progeny of each stand were planted per replication 

 as a four -tree row plot. 



Data were taken by estimating the number of dead or 

 dying (off-color) branch or leader shoots. The range of 

 damaged shoots was determined by sample counts of 

 damaged shoots throughout the test on several trees. 

 From this inspection a scoring system was developed to 

 apply to all trees. Then the amount of infestation for each 

 tree was estimated and placed in one of the following 

 classes: 







= 



infested branch or leader shoots, 



mean 



= 



1 



= <5 



infested branch or leader shoots, 



mean 



= 2.5 



2 



= 6-32 



infested branch or leader shoots, 



mean 



= 19 



3 



= 33-67 



infested branch or leader shoots, 



mean 



= 50 



4 



= 68-100 



infested branch or leader shoots, 



mean 



= 84 



Tree height varied from 3.6 ft (1.1 m) to 30.0 ft (7.0 m), 

 and so to determine the proportion of shoots damaged on 

 a tree basis, a regression of the total number of tips per 

 tree (healthy and damaged) by height was determined. A 

 random sample of 224 trees (6 percent of total trees) was 

 selected to develop the following regression formula: 



estimated percent damage = number damaged shoots/ 

 TT where TT = a + bx, and TT = total shoots, x = tree 

 height, a and b = regression coefficients, number dam- 

 aged = the mean of the estimated number of damaged 

 shoots. 



A goodness-of-fit procedure was used to compare actual 

 shoot damage with estimated shoot damage. Because 

 estimated damage covered a wide range, it was trans- 

 formed (arcsin V percent damage) (Steel and Torrie 1960). 



1 



