Marbled Murrelets nest secretively high in the canopies of very large old-growth trees up to 50 km from the sea. They fly to and from their inland nest sites in the semi-dark before dawn and after dusk. These behaviours make it difficult to find nests and census the birds. High-frequency radar (the type used on ocean-going vessels) has proved to be extremely useful for detecting and counting murrelets as they fly from the sea to their nest sites.
We use a Furuno FR-810D 10 kW marine surveillance unit using 9410 MHz
(X-band) transmitted through a 2 m scanner. The radar scanner can be
mounted on a shore platform, the top of a truck or on a small vessel
and positioned where murrelets can be most easily detected. Power is
provided by 12V deep-cycle batteries.
The radar is usually located near the shore where the murrelets cross from the sea to fly inland. The best results are obtained in narrow fjords or valleys where the flight-path of the murrelets is narrow and crosses within the scanning radius of the radar.
Repeated radar counts made at the same site show considerable seasonal variation and also some variation from year to year. Multi-year sampling is therefore advisable and on Vancouver Island we get least variability by restricting our counts from mid-May through mid-July.
Our multi-year research in Clayoquot Sound provided reliable estimates of the numbers of murrelets using the 18 major watersheds in the area. These could then be ranked for their importance to murrelets, ranging from Watta South with an average of 45 birds up to Moheya (596 birds) and Watta (650 birds).
Our UVic murrelet group is currently doing similar surveys in 30 watersheds on southwest Vancouver Island, around Barkley Sound. Here are some preliminary results - the size of the circle shows the number of murrelets detected with radar.
By comparing the numbers of murrelets counted with radar with various habitat measures within each watershed we have identified some key features affecting murrelets at the large, landscape-level. In Clayoquot Sound counts of murrelets were strongly correlated with areas of remaining old-growth forest containing large trees and in valleys below 600 m elevation (see Burger 2001). Murrelet numbers were also correlated with the absolute size of the watershed (big valleys supported more murrelets as expected), and showed a negative effect of the area of old forest recently logged. Other radar studies in British Columbia (Burger et al. 2004) and on the Olympic Peninsula (Raphael et al. 2002) have found similar trends.
The strong linear correlation between murrelet numbers and areas of remaining old-growth forest suggest that murrelet numbers would decline as the area of available suitable forest was reduced by clearcut logging. This was clearly shown in our Clayoquot Sound data where watersheds with no logging or minor amounts were strongly correlated with the areas of forest originally found in each watershed, but three of the five watersheds which had been heavily logged had fewer murrelets than expected (see Burger 2001 for details).
In the graph each point shows the data from a watershed in Clayoquot Sound. The white circles are watersheds with no logging, yellow triangles are watersheds with minor amounts (<10% area affected) and the red squares are heavily logged valleys (30-50% logged).
The significant relationships between numbers of murrelets counted with radar and the areas of large-tree old-growth within each watershed allows estimates of density to be made. Pooled data from 5 independent studies indicate that densities of murrelets on western Vancouver Island are significantly higher than those on the mainland of British Columbia (Burger et al. 2004). The causes for these differences are not clear but might reflect the quality of both marine and inland forest habitats. Relatively few murrelets nest at the heads of the long fjords on the BC mainland even though there is suitable forest habitat there.
Radar is now widely used as a census and monitoring tool for murrelets across British Columbia. A standardized protocol has been developed and there are now over 170 sites which have been surveyed for murrelets with radar. Analysis of these data by Dr Peter Arcese (UBC) and his collaborators shows that radar counts provide a reliable method for detecting changes in murrelet populations within regions (Arcese et al. 2005).
Samples of our radar studies:
Burger, A. E. 1997. Behavior and numbers of Marbled Murrelets measured with radar. Journal of Field Ornithology 68:208-223. [pdf version]
Burger, A. E. 2001. Using radar to estimate populations and assess habitat associations of Marbled Murrelets. Journal of Wildlife Management 65:696-715. [pdf version]
Burger, A. E. 2002. Radar inventory and watershed-level habitat associations of Marbled Murrelets in Clayoquot Sound, 1996-1998. Pp. 35-56 in Multi-scale studies of populations, distribution and habitat associations of Marbled Murrelets in Clayoquot Sound, British Columbia (A. E. Burger and T. A. Chatwin, eds.). Ministry of Water, Land and Air Protection, Victoria, BC. Available at: http://env.gov.bc.ca/wld/documents/techpub/mamuwebs.pdf
Burger, A. E., T. A. Chatwin, S. A. Cullen, N. P. Holmes, I. A. Manley, M. H. Mather, B. K. Schroeder, J. D. Steventon, J. E. Duncan, P. Arcese, and E. Selak. 2004. Application of radar surveys in the management of nesting habitat of Marbled Murrelets Brachyramphus marmoratus. Marine Ornithology 32:1-11. [pdf version]
Ronconi, R.A., C.C. St. Clair, P. D. O'Hara, and A.E. Burger. 2004. Waterbird deterrence at oil spills and other hazardous sites: potential applications of a radar-activated on-demand deterrence system. Marine Ornithology 32: 25-33. [pdf version]
Extension note and other radar articles: