tillers in samples comprised of five adjacent rice plant hills.
Twenty such samples were checked in each plot. Sample locations were
marked so posttreatment appraisals could be repeated on the same
plants.

At the end of the pretreatment evaluation period, two of the sample
plots were selected for treatment and 120 stakes were placed in each.
These stakes were coated with a "bird glue" (prepared from sap of
local trees) and placed in the paddies so that the coated ends pro-
truded 15-20 cm above the rice plants. The two remaining plots were
left untreated and served as reference areas. Bird counts and damage
appraisals were then repeated as described above. The resultant data
were subsequently analyzed by a two-factor analysis of variance.

The average number of birds per 10-min interval were similar (11.6 in
treated plots and 7.9 in reference plots) prior to placement of the
glue stakes. After treatment, bird numbers dropped to as low as 0.2
bird per interval in treated areas and increased to as high as 17.1
birds per interval in untreated areas. During the pretreatment sur-
vey, damage averaged 14.5 percent in test plots and 11.8 percent in
reference plots. Posttreatment surveys revealed increased damage in
all plots. These increases were slight (up to 19.9 percent) and not
significant (p <0.005) in treated plots, but quite large (up to 63.2
percent) in reference plots. Although the stakes lost their adhesive
quality within 3 days, Lonchura learned to avoid plots containing them
for a full 8 days after treatment.

Use of crown baiting with wax blocks to reduce rat damage in coconut
palms
Crown baiting with anticoagulant rodenticides has been proven effec-
tive in reducing rodent damage to coconuts. However, ease of place-
ment of bait in the crowns and deterioration of loose bait materials
are among the problems encountered in the use of this technique.
Incorporation of the bait into wax blocks might eliminate some of
these difficulties.

Two 1-ha study plots were selected within a large area of coconut
plantations near Victoria, Mindoro Oriental. The plots were separated
by over 200 m, and each contained 100 palms of a local variety. One
plot was selected for treatment, the other served as a reference. For
treatment, 25 randomly chosen palms were baited monthly with paraffin-
ized cubes containing polished rice, talc, and warfarin (0.025 per-
cent). The cubes were placed in the palm crowns with a long bamboo
pole. Monthly records of bait additions, rat activity (tracking
tiles), damaged nuts, and coconut production were kept for 1 year.