It would be nice to have all the contours digitized.  There are, however, other ways of looking at the same information.  Imagine that you have the points loaded with weights that reflect all the real-world information you have including info from contours.  Here is a sample of ways one can look at "loaded" points--the points pull the surface, with a sort of tension that produces peaks and pits based on the loadings of the sample points.  Try some of the early points, loaded, and see if later sites fell on or near peaks of the loaded surface.  Then try again with all sites...

I assigned arbitrary weights to your points; then, used Spatial Analyst to contour the distribution.  I chose to use "spline" to use "weight2" and to use "tension" when asked various questions.  Splines are smooth curves formed from piecing together cubics in such a way that where cubics of different equations come together, the slope of the tangent line from the left and the slope of the tangent line from the right are the same.

A line pattern like the one above can be difficult to use as a backdrop--it's too busy for that.  However, the same pattern can be displayed in a continuous range of color that is suitable for backdrop, as below.

There are packages that let you easily calculate "home ranges" (one from the USGS and the Alaska Biological Station)--here is one based on a "kernel"--it pulls out the two separated parts and surrounds the kernel with probability ellipses.


 

Or, one can just do the same sort of thing for the whole distribution of sites; here the largest ellipse is a 95% probability ellipse; the smallest is 5%.