I saw this posted to PABirds recently and got permission from Scott to cross-post as I think it's relevant to people searching for N. Saw-whet Owls, particularly during CBC season.
 
Tyler Bell

California, Maryland 

Subject: Re: Fw: saw-whets & long-ears
From: Scott Weidensaul 
Date: Thu, 3 Dec 2009 10:08:28 -0500

  For what it's worth, we've radio-tagged more than 50 saw-whets in 
the past eight years to study both their roost selection and (for the 
past two years) their nocturnal activity ranges. Most of this work 
was done in Michaux State Forest in Cumberland County, as well as 
some tracking in Weiser State Forest in Schuylkill and Dauphin 
counties.

  Before the leaves drop, saw-whets often roost in hardwoods, usually 
high in the canopy and toward the outer edge of the tree, tucked up 
in a clump of leaves. If you think warbler neck is bad, "saw-whet 
neck" is worse, trying to locate a small owl high in an oak or maple. 
After leaf-fall, they move to whatever thick, evergreen cover is most 
abundant. In Michaux SF, it's almost always pitch pine, with white 
pine, mountain laurel and rhododendron also used on occasion. The 
handful of owls we tracked in Weiser SF used the same species, plus 
hemlock, which is more abundant there.

  They change roosts every couple of days, sometimes daily, but last 
year we found several owls alternating in the same couple of roost 
trees, which to our eyes looked identical to ever other pitch pine in 
the woods, but which obviously offered something special for an owl. 
The same owl may roost 70 feet up in a pitch pine one day, and 
literally on the ground in a mountain laurel thicket the next.

  Interestingly, we've found that the saw-whets rarely hunt in the 
same areas where they roost, so I'm not sure that searching for roost 
sites immediately adjacent to high rodent populations is a good 
strategy. In some extreme cases, we've had saw-whets fly nearly three 
miles from their roost on the top of South Mountain, down into the 
valley to hunt a small farm woodlot, then back up to the mountaintop 
again before morning. Most don't travel that far (usually a quarter 
to half a mile), but we've never had one stick around the roost site 
to hunt.

  Another interesting tidbit our telemetry crews have discovered is 
that saw-whets are very active hunters. I always assumed they sat and 
waited for long periods, but for the first couple of hours after 
dark, and again in the last couple of hours before dawn, they are 
constantly on the move. I now suspect they hunt more like accipiters, 
flying short distances, perching and scanning for prey, then moving 
on again after a minute or two until they make a kill.

  All this information about nighttime activity came at great expense 
in terms of time and effort by our '08 and '09 telemetry crews - Drew 
Weber, who is heading the team this year as research tech (and who 
was last year's intern), our '08 tech Anna Fasoli, and this year's 
interns, Hannah Panci of Wisconsin and Kim Romano of Georgia. They 
have spent countless nights in all kinds of weather radio-tracking 
these owls and uncovering all this new information about them. Aura 
Stauffer from DCNR has been heading up the roost site tracking since 
2004, is responsible for most of the information we've gleaned about 
that aspect of saw-whet ecology.

  As for long-eared owls, I do suspect that proximity to good hunting 
habitat (open grasslands and fields) is a factor in which roost sites 
are chosen. I'd also add that researchers in the West have found that 
longears like very, very thick cover for their roosts, and that by 
the time a thicket has grown up enough for humans to squeeze into, it 
may no longer be most useful for the owls.

  Sorry for the long post,

  Scott Weidensaul
  Schuylkill Haven, PA