The increase in growth when phosphorus alone was applied indicates P
was a limiting factor and that a fair amount of nitrogen was being sup-
plied by the soil. Variability caused least significant differences
(LSD's) to be high and only two nitrogen source, urea and Osmocote 3,
produced significant increases in growth when compared to the phos-
phorus only treatment. Nitrogen sources, which were not effective,
were sodium nitrate and coarse-textured IBDU.
Results were also variable for S. cynosuroides (Table 4). The best
N sources were Osmocote 8-9, Mag Amp + Osmocote 3, and Urea. The re-
sults for both species were inconclusive due to high variability; how-
ever, it was apparent that phosphorus and some form of nitrogen were
necessary for good growth.
Conventional materials, ammonium sulfate and concentrated super
phosphate appear to be the most practical sources of nitrogen and
phosphorus.
Effects of potassium.   Potassium applied without N and P had no
effect on growth of S. alterniflora. Added N and P increased growth 20
times over the unfertilized check; however, an additional 112 or 224
kg/ha potassium did not result in additional increased growth. When the
rate of K was increased to 336 kg/ha there was a significant increase in
dry weight.
Direct seeding of Spartina does not appear to be nearly as promis-
ing as might be expected. Although wave action, the major hazard to
seedling survival in many astronomic tide situations, is absent here,
erratic water levels restrict survival. Extended periods of high water
eliminate many seedlings in the lower range while like periods of low
water, accompanied by low rainfall and high temperature, drastically


14