1981. The pit was created over 30 years 

 ago as a source of fill material for the 

 overseas highway. The dredged site is 1.3 

 to 1.7 m (4.3 to 5.6 ft) deep and is cov- 

 ered with fine calcareous sand and silt. 

 The surrounding area is 0.3 to 0.7 n (1 to 

 2 ft) deep and is well vegetated, primar- 

 ily with turtle grass, and portions of the 

 borrow pit were gradually being revege- 

 tated. 



The experimental design used a total 

 of 22 combinations of plant species and 

 transplantation techniques. Bare single 

 short shoots and plugs of seagrass plus 

 sediment (22 x 22 x 10 cm) were used for 

 turtle grass, manatee grass, and shoal 

 grass. Seeds and seedlings of laboratory- 

 raised and field-collected turtle grass 

 were planted, but seeds and seedlings of 

 the other species proved impossible to 

 find in sufficient quantity. Short shoots 

 were attached to small concrete anchors 

 with rubber bands and placed in hand-dug 

 holes 1 to 3 cm deep, which were then 

 filled with sediment. Seeds and seedlings 

 were planted by hand without anchors after 

 it was determined that anchors were 

 detrimental to the survival of the seed- 

 lings. The large sediment plugs with 

 seagrass were placed in similar sized 

 holes made with another plugging device. 

 Plugs and short shoots of all species were 

 planted with both 1- and 2-m spacing, 

 while the seeds and seedlings of turtle 

 grass were planted using 0.3-, 1-, and 2-m 

 spacings. 



Of the 20 manipulations of species, 

 planting techniques, and spacings, only 

 three groups survived in significant num- 

 bers for the full 2 years: manatee grass 

 plugs with 1-m spacing, and turtle grass 

 plugs with both 1- and 2-m spacing. Tur- 

 tle grass plugs showed the hiqhest sur- 

 vival rate (90% to 98%), but did not 

 expand much, increasing their coverage by 

 a factor of only 1.6 during the 2 years. 

 Manatee grass spread rapidly from plugs 

 under the prevailing conditions and had 

 increased its area by a factor of 11.4 in 

 the 2-year period. The initial planting 

 of shoal grass, however, died out com- 

 pletely after only a few months, and a 

 second planting was made with larger, more 

 robust plants from a different site. This 

 planting survived sufficiently to increase 

 its area by a factor of 3.4 after 1 year. 



The transplants using short shoots of 

 the various species were not nearly as suc- 

 cessful. Although some of the treatments 

 showed short-term growth and survival, 

 none of the treatments using short shoots 

 survived in significant quantitites. Sim- 

 ilarly, the freshly collected seeds and 

 seedlings of turtle grass showed no long- 

 term survival at the barren transplant 

 site, and showed only 4% survival when 

 planted into an existing shoal grass bed. 

 Seeds and seedlings that had been raised 

 in the laboratory showed a modest survival 

 of 29% when transplanted to the field, but 

 even the survivors did not spread signifi- 

 cantly. 



Although several of the restoration 

 techniques used by Lewis et al . (1981) 

 proved to be technologically feasible, 

 there are still major logistic and eco- 

 nomic problems remaining. The plug tech- 

 nique showed the highest survival rate, 

 but the cost estimates ranged from $27,000 

 to 86,500/ha. Because of the large volume 

 and weight of the plugs, this method 

 requires that large source beds be close 

 to the transplantation site. The removal 

 of large quantities of plugs can represent 

 a major source of disturbance for the 

 source bed, as the plug holes are as slow 

 to recolonize naturally as propeller cuts 

 and other similar disturbances. Despite 

 the spreading recorded at the transplant 

 site, the source holes for the plugs did 

 not show any recolonization at the end of 

 the 2-year period. If source material was 

 required for a large scale revegetation 

 project, the disturbance caused by the 

 acquisition of the plugs could be a major 

 impact itself. For this reason Lewis 

 et al . (1981) suggested that this method 

 be mainly used where there are source beds 

 that are slated for destruction because of 

 some developmental activity. 



The only other technioue that showed 

 any significant survival was the utili- 

 zation of laboratory cultivated seeds 

 and seedlings. This method was prohibi- 

 tively expensive with costs estimated 

 at $182,900/ha, largely due to cultiva- 

 tion costs; survival was still below 

 30%. Seeds and seedlings are also suit- 

 able only in areas where the water motion 

 is relatively quiescent, as their abil- 

 ity to remain rooted at this stage is 

 minimal . 



92 



