Precipitation was found to have little direct affect on water tempera- 
ture. Precipitation modified maximum and minimum temperatures during, 
and for a period of 24 to 48 hours following, the storm, but the 
variations were small and did not lower average water temperature. 
Similar results were noted at the Coweeta hydrologic laboratory (3). 
The factor which probably exerts the greatest modifying influence is not 
precipitation, but rather the overcast skies and cool air temperatures 
associated with the storm. These factors are jointly related and it is 
difficult to test the influence of any one of them. 
STREAM CHANNEL CHANGE 
Continuous changes in beds and channels of streams are natural pheno- 
mena. Most of the valleys and mountains of both the mainland and the 
islands were buried under an ice sheet during the Pleistocene epoch. 
Glaciation has resulted in the formation of broad, flat-floored and U- 
shaped valleys with steep sides (2). Earthslides occur as a result of 
weathering of certain rock formations and saturation of soils which 
allows the mass to slide from the steep slopes. Aerial photos and ground 
reconnaissance indicate that approximately 250 years ago a large slide 
blocked Harris River and changed its course, 
The choking of stream channels by material other than landslides is of 
relatively common occurrence. High flood water undermines many trees 
growing along the stream channel. Stream banks are cut in one section 
and filled in another; new gravel bars are formed, others are washed 
out. Evidence of this can be readily seen in mudbars, cutbanks, in log 
tangles embedded in stagnant pools, and in many other forms. Especially 
vulnerable to alteration are unstable streams which are subjected to 
flash floods during periods of heavy precipitition. This change is 
characteristic of many of the salmon streams of Southeast Alaska. 
The objective of this phase of the study is to determine to what extent, 
if any, the process of stream channel change is accelerated by logging. 
Debris, Windfalls, and log Jams 
Table 10 shows the number of branches, broken logs and/or trees pro- 
truding into the main channels of the study streams during the pre-logging 
period 1949 to 1953, inclusive. The period has been one of accumulation 
with Maybeso Creek showing an average yearly accumulation of 30 pieces 
per mile, Harris River 19 pieces, Indian Creek 20 pieces, and Old Tom 
Creek, 33 pieces. These values indicate that a considerable amount 
of debris exists in all study streams under natural conditions. 
In what form has this accumulation taken place? Is the debris accumula- 
ting as scattered material, or as log jams of constantly increasing 
size? Table 11 shows jams by total number and size. A large number of 
jams does not exist in any of the streams. The most rapid build-up of 
jams has occurred in Maybeso Creek with an accumulation of 4 jams con- 
taining 10 to 14 logs, 2 containing 15 to 29 logs, and 1 containing 30 
=- 30 - 
