Dear ZOOARCH:  Alexander Fenton (1978, 448) in his study of the lives and organisation of crofting communities in the time and place where they last flourished (the Scottish Isles at the turn of the previous century), noted small tenant farms had thrity to forty sheep.  He lists the 1902 sheep census for 1902 North Ronaldsay (p. 470), which lists the holding by croft (so you could get means and s.d.s; its possible that some larger flocks are sheep-raising specialists' rtaher than mixed farming-fishing households).  He presents the contemporary Sheep management regulations in full, for their information on communal organisation of the task (p. 471 - 2), and the various sheep-markings used (p. 472-490).  The chapter also has lots of other info on longevity, number of offspring meat and wool yield, and proportions by age and sex.

 

Fenton, A. (1978):  The Northern Isles: Orkney and Shetland (Edinburgh: John Donald)

 

And I would bet that there will be no way to determine whether an archaeological assemblage represents the number of sheep for a given croft.  How often does a whole croft (or any other economic unit, like a Roman villa or aisled hall) get excavated, giving us access to all its bones? How often do all the bones get retrieved? How do we tell which are so closely contemporary that they can be a snapshot of a population?  Using faunal remains as a sample (following a little thought about the nature of retrieval and preservation bias) is fine.  Since we don't have all the bones, deriving meat yield or calories and dividing by 2000 calories per day does not really give us any reliable picture about the human population.

 

Greg Campbell