It nests in old-growth and older-aged trees (up to 60 km inland) and on the ground (only in northerm part of range) (Piatt and Ford 1993, Gaston and Jones 1998, Burger 2002, McShane et al. 2004, Piatt et al. 2006, Barbaree et al. 2014), with the breeding season stretching between March and September in California, April and September in British Columbia, and May and September in Alaska (Piatt et al. 2006). Forest areas with multiple canopy layers and high moss abundance are strongly preferred. Research in British Columbia shows that in areas where forest habitat is relatively plentiful the species seldom re-use the same trees as nest sites, whereas in areas where logging has reduced old-growth there is a higher proportion of nest tree re-use (Burger et al. 2009). The species has been shown to suffer increased predation and disturbance at forest edges adjacent to recently cleared areas, compared with forest edges adjacent to regenerating or riparian areas (Malt and Lank 2009). Multiple radar and modelling studies have shown a significant correlation between numbers of birds entering watersheds and the areas of suitable forest habitat within those watersheds (Burger 2002, Burger et al. 2004, Raphael et al. 2011). Breeding is mid-March to early September in California, but more compressed further north (Hamer and Nelson 1995a, b, Gaston and Jones 1998, Burger 2002). The diet is sandlance, herring, other small schooling fish and, in winter, invertebrates (Gaston and Jones 1998, Piatt et al. 2006). Chicks are generally fed large subadult or adult prey rather than juveniles or larvae (Piatt et al. 2006). It feeds in near-shore habitats up to 1.4 km offshore, in sheltered waters, lagoons and sometimes inland lakes (Carter 1986, Hunt 1995, Gaston and Jones 1998, Burger 2002, Hebert and Golightly 2008). The effect of marine attributes on murrelet productivity varies from increasing density and productivity estimates with increasing sea surface temperature in the San Juan Archipelago, Washington (1995-2007; Raphael and Bloxton 2008), to decreasing reproductive efforts with warmer sea-surface temperatures in California (Bigger and Chinnici 2008). Sea-surface temperatures and other marine attributes explained little of the variation in murrelet distribution and abundance compared to the amount and cohesion of forest nesting habitat (Raphael et al. 2015). Daily movements to feeding areas can be up to 250 km in exceptional cases (Whitworth et al. 2000), but are normally about 30 km (Burger 2002, Piatt et al. 2006). Radio-marked birds from Redwood Creek in North California moved a maximum average distance of 99 km alongshore, with males travelling further than females, and non-breeding males travelling further than breeding males, perhaps in search of mates or nesting habitats. Average home range size was 505 km², again being greater for males than females (Hebert and Golightly 2008). Individuals exhibit plasticity in their foraging behaviour, foraging closer to shore and increasing dive rates during nesting (Peery et al. 2009). Marbled Murrelets most often forage in pairs (Piatt et al. 2006). Individuals in the northern part of its range may travel south during the non-breeding season, a movement which likely reflects the availability of prey (Piatt et al. 2006).

Brachyramphus marmoratus occurs in the U.S.A. and Canada in California, Oregon, Washington, British Columbia, south-east Alaska, Prince William Sound, Kenai Peninsula, Lower Cook Inlet, Barren Islands, Afognak and Kodiak Islands, the Alaska Peninsula and the Aleutians locally to Andreanof and Near Islands (Gaston and Jones 1998). In Alaska (70% of the population), historical estimates place the population at c. 750,000 individuals, though when trend estimates are applied to this figure it gives an estimated 2006 population of c. 271,000 individuals (Piatt et al. 2007). The British Columbia population was previously thought to be c. 54,000 - 92,500 (Piatt et al. 2006) but extrapolations from recent radar counts suggest the population may in fact be c. 72,600-125,600 birds (COSEWIC 2012). This higher estimate is likely due to differences in survey methodology as opposed to a genuine population increase. The population in Washington, Oregon and California is estimated at 15,400-23,900 individuals (Falxa et al. 2014, Falxa and Raphael 2016). The greatest historical decreases have occurred in Washington, Oregon and California (A. Burger in litt. 2012, Falxa and Raphael 2016). Declines are also reported in British Columbia and south-east Alaska (Perry 1995). Trend analyses conducted during 2000-2013 suggest a decline of c.11% over the period in Washington, Oregon and California (Falxa et al. 2014, Falxa and Raphael 2016), and a decrease of c.70% in Alaska from the 1980s to 2006 (Piatt et al. 2006). At-sea surveys over the past 25 years in British Columbia suggest declines of c.1% per year (Piatt et al. 2006), although radar surveys suggest the population may have been relatively stable since 1999 with declines in some regions of British Columbia (COSEWIC 2012, Bertram et al. 2015). Availability of nesting habitat in British Columbia, which is strongly correlated with local breeding populations (Burger 2001, Burger et al. 2004), has declined by 22%  between  1978 and 2008, and is continuing (COSEWIC 2012). 

Conservation Actions Underway
It is Threatened in all range states and provinces except Alaska. Detailed conservation recommendations were made in 1990s (USFWS 1992, Ralph et al. 1995) with a Habitat Conservation Plan drafted in 1997 for Washington State (Wilson 2019). Federal land-use in the USA is regulated, areas for management identified and surveyed, and some temporarily removed from logging (Nelson 1997, Raphael 2006). In August 2016, the USFWS released the final determination on critical habitat for the species in the listed range and in January 2010 determined that the species still warrented protection and listing as Threatened after numerous challenges by the timber industry. USFWS initiated a status review of the species in 2008, which will also function as a 5-year status review. In Canada there has been extensive research, an updated recovery strategy (Environment Canada 2014), some (relatively minor) habitat protection under the British Columbia Forest and Range Practices Act, more extensive protection of forest habitat under various Land Use agreements and a radar monitoring plan developed by the Canadian Marbled Murrelet Recovery Team (CMMRT 2003, COSEWIC 2012). The Canadian Marbled Murrelet Recovery Team developed a Recovery Strategy to be compliant with the Canadian Species at Risk Act (Environment Canada 2014). This lays out the general strategy for population stability (population reduction between 2002 and 2032 not to exceed 30% of the 2002 population and this decline is linked to similar limited decline in available suitable nesting habitat). The essential Marine and Terrestrial Action Plans required by the Species at Risk Act have not been drafted (as of 2016) and these are not likely to be completed and implemented before 2020. Under the Canadian Recovery Strategy critical nesting habitat requiring protection is defined as 70% of the 2002 suitable nesting habitat coast-wide, with varying percentages (68-90%) in the six conservation regions in British Columbia (Environment Canada 2014).

The U.S. Department of Natural Resources (WDNR) began developing a Marbled Murrelet Long-Term Conservation Strategy in 2007 (Escene 2007); the final strategy still has not been released as of 2016. The Northwest Forest Plan (1994) is expected to ensure the protection of a large proportion of important habitats in the USA (Raphael 2006).

Extensive areas of suitable forest nesting habitat have been set aside in conservancies on the northern and central mainland and in Haida Gwaii (formerly Queen Charlotte Islands) (COSEWIC 2012). Smaller areas are being protected by other forestry and conservation measures. Overall, an estimated 35% of the 1,826,828 hectares of  suitable habitat in all of British Columbia (based on the Canadian Marbled Murrelet Recovery Team modeling criteria [CMMRT 2003]) had been protected under various measures by 2011 (COSEWIC 2012). In 1998, the Exxon Valdez Trustee Council protected 179 km² of Afognak Island (BBC Wildlife 1999 172: 23). In 2007, 1,569 ha of forested land on the Oregon Coast was acquired under conservation easement for the species (amongst others), part funded by the New Carissa oil-spill funds (Escene 2007). Between 1998 and 2002, 507 Marbled Murrelets were radio-tracked in British Columbia (Zharikov et al. 2007), during 2005-2007, 111 birds were radio-tracked at Port Snettisham, Alaska (Nelson et al. 2008), and between 2004-2008, 157 birds were radio-tagged on Olympic Peninsula, Washington to determine nesting habitat, activity patterns and distribution. A recommended protocol for surveying the species in forests was published in 2003 by PSG (Mack et al. 2003). In British Columbia standard protocols have been developed for various survey methods (RISC 2001, Manley 2006). Research has shown that habitat management at relatively fine scales may provide conservation benefits (Horton 2008) and that the species would benefit from a reduction in the amount of hard edges (recent clear-cuts) at both patch and landscape scales (Malt and Lank 2009).

Conservation Actions Proposed
Survey potential nesting habitat on all ownerships and protect all occupied sites. Collect data on the ratio of juvenile to adult birds from sites throughout the range, and monitor over time, as this is thought to be a reliable indicator of productivity (Peery et al. 2007). Research means of improving the abundance of high quality food, e.g. small fish, during the pre-breeding period. Minimise damage to fish stocks and feeding areas (RENEW Report 1999 9: 20). Conduct research on the behaviour of migrants to determine the extent to which dispersal results in gene flow and prevents declines in resident populations (Peery et al. 2010). Complete and implement the SARA-compliant Canadian Marbled Murrelet Recovery Action Plans needed to implement the Canadian Recovery Strategy (Environment Canada 2014). Protect both terrestrial nesting habitat and marine habitat throughout the species's range (Lorenz et al. 2016), create larger blocks of habitat, provide buffers and other means for minimizing edge effects and predation (K. J. Kuletz in litt. 1999, USFWS 2012). Renew the radar monitoring at British Columbia watersheds (Bertram et al. 2015), in order to track population trends and determine the effects of habitat removal (logging) and local changes in marine conditions. Move campgrounds away from old-growth areas in Californian State Parks, in order to reduce predator populations in breeding areas. Reduce oil-spills, gill-net mortality and logging (K. J. Kuletz in litt. 1999). List as Threatened in Alaska (K. J. Kuletz in litt. 1999).