
covering a greater sample af the total woodcock breeding range, run 
only once, might result in less sampling error than repeating counts 
over the same route if this necessitates covering less territory. 
It is becoming apparent that the main weakness in our present 
effort to obtain a breeding ground index of woodcock abundance is that 
we are attempting to measure populations in areas of suitable habitat 
but without consiceration of how much change is taking place in the 
total amount of this habitat. Our routes are laid out in such a way 
as to cover the largest amount of good woodcock habitat possible in 
the time available. In fact we have recommended that routes be changed 
when ecological succession advances to a stage unsatisfactory for wood- 
cock breeding. This may turn out to be a reliable way of measuring 
woodcock population changes within the favoreble habitats, but this is 
only half the job. It in no way measures the changing amounts of favor- 
able habitat available from year to year. Studies are in progress under 
the auspices of the game departments of the states of Michigan, 
Pennsylvania and New York, and under the supervision cf the Massachusetts 
and Maine Cooperative Wildlife Research Units to determine what consti- 
tutes habitat suitable for supporting breeding woodcock. The next step 
in improving our present system of appraising over-all changes in 
woodcock populations world seem to be to devise a way of determining the 
annual change in amount of suitable woodcock habitat. If we could obtain 
such an index of abundance of favorable habitat, together with an index 
of abundance of woodcock within the favorable habitat, we would have 
everything we need. The only alternative would seem to be to have a 
large number of routes distributed completely at random throughout the 
breeding range of the woodcock without any consideration being given to 
the occurrence of suitable habitat, In all probability in most areas 
this method would result in many routes being run without recording any 
woodcock, The most difficult problem would probably be to maintain co- 
operator interest sufficiently high to get enough routes run m this 
basis. 
During 195 a total of 193 routes were run in 17 states of the 
United States and 21 counts in 5 provinces of Canada. These counts 
which are summarized in the following papers were mace possible by the 
cooperation of the game departments of the various states and provinces, 
wildlife research units at land-grant colleges in Maine, Massachusetts 
and Pennsylvania, the Wildlife Management Institute and the Royal 
Mounted Police in Canada, as well as many individuals, 
The combined data for each state are presented in Table l. 
