(4) Is the site representative of a type of habitat 

 which is rapidly being converted to other land 

 uses? 



(5) Would loss of the habitat constitute an irretriev- 

 able loss of resources to Hyde County? 



The inventory results reflect a bias toward large areas of 

 relatively undisturbed land. A chief limit inherent in the 

 study is that it was too broad; more attention should have 

 been focused on analysis of communities at specific locations. 

 While such an approach would have satisfied the desire for 

 technically complete community descriptions, it would have 

 diverted us from our objective to present useful natural 

 areas data in the context of the county and its land use 

 patterns as a whole. We recognize that certain biologically 

 significant areas - and significant features at identified 

 sites - have gone unnoted and unreported by us. We wish to 

 point out the following areas in need of further inventory: 



(1) Brackish marshes in the upper portion of the 

 Pungo River 



(2) Swamp forest along New Lake Fork of the 

 Alligator River 



(3) Additional survey work in the New Lake Fork Pocosin 



(4) further survey work on Roper Island 



(5) Mesic hardwood flats east and southeast of New Holland 



(6) additional survey work on low pocosin area of Gull 

 Rock Game Lands 



(7) aquatic and marsh communities around Lake Matta- 

 muskeet 



(8) further survey work in the wilderness portions of 

 the Swanquarter National Wildlife Refuge 



(9) further survey work in pond pine pocosins along 

 the Dare County line. 



(10) the "bird refuge" tract of the Swanquarter National 

 Wildlife Refuge. 



12 



