Trailing edges

As climate change impacts ecosystems, we expect to find two different "edges" to species' ranges. Where the climate is making new places habitable for a species, we have a "leading edge" where the species range is expanding. For example, the southern edge of where Saltwater Crocodiles live is going to be the leading edge for that species, assuming that warmer conditions in the ocean allow salties to expand into more southern areas in Australia.

Salties surfing south on the "leading edge" of their habitat envelope.

But there may also be an edge where climate change is making conditions less hospitable for a species, so that the species may be locally eradicated in these areas. The lower elevation extent of alpine species ranges, which is moving upslope in many cases, is an example of this "trailing edge".

But what actually happens at the trailing edge - how do climatic changes interact with the needs of a species to make their current range inhospitable? This is an interesting question, although a fairly bleak one in many ways - and is important. How slowly or rapidly ranges contract at the trailing edge will determine a lot about the risks of population declines or even extinction as the climate changes. And this is what we explored in a recent publication brilliantly designed and led by Steph Carlson at Berkeley, which you should totally read here.

Salmonid fish in California have always been at the "difficult" edge of salmon habitat. That's because salmonids need cold water conditions to thrive, becoming metabolically stressed as temperatures in the water increase above about 19 oC. This actually makes them super valuable for salmon populations overall, as they are better adapted to warmer waters than their more northerly cousins - and those genes could be really useful to help the broader salmonid population adapt to overall warming waters.

Chinook Salmon run, Northern California

The other thing that salmonids need, famously, is access to their breeding habitat, often a long way up rivers in headwater areas. And it was access to breeding pools, rather than warm water, that caused the trouble that Steph Carlson and her colleagues (including me!) studied during the 2011-2014 California drought.

The drought caused really unusually low flow conditions in the rivers (this is the bit I helped with). These are shown in the plots below. The panel called "A" shows the usual flow in the winter period (Nov - Jan) over 40 years. Look for the circles - that's where 2013 flows were. In all but one of the rivers in Northern California this was the lowest flow winter in those 40 years. The flows also arrived really late. This is shown in Panel B. This plots the date when the winter flow regime started. For all but 2 of the rivers, the start of winter flows in 2013 was the latest on record, and in all rivers that start ranged from the end of Jan to the start of February.

The late arrival of winter flows meant that many of the salmon species in the Northern California rivers arrived at their breeding pools really late compared to usual conditions. This is shown in the plots below - the 2014 arrivals are shown in colour compared to normal conditions shown in grey. While the delay varied with location and species, many locations that would normally have seen spawning salmon arrive in November - December didn't experience that migration until March.

The flow conditions also impacted where the fish were breeding and spawning. This is shown in colour for the adult fish in the Mattole River, and for baby fish in the Russian river. The Mattole River has an intermittent bar that was closed for most of the winter, and a large boulder that blocked the fish from getting past it under the low flow conditions. This is why 100s of adult fish were "stuck" close to the coast in the Mattole in 2013, compared to being spread throughout the whole river in other years (red colours in panel A). Baby fish didn't occupy anywhere near as much of their habitat as usual, with tributaries containing juveniles reducing from 19-20 in normal years in the Russian River to only 4 in 2013-2014 (blue in panel B).

The consequence of these changes, as determined by Steph and her fish ecology friends, is that the 2013, 2014 and 2016-2017 breeding cohorts of Coho Salmon in the Russian River was effectively lost. In fact, the Russian River Coho population was only rescued by the releases from a captive breeding program, which exists because of the vulnerability of this salmonid population to extinction. However, the different species experienced different impacts - which were severe for Coho, bad for Chinook, and relatively mild for Steelhead.

The delayed rains affected 3 species ...—but the impacts differed across species due to differences in their breeding phenology... [some experienced] contracted breeding and rearing distributions; reduced juvenile production; and, in three extreme cases, complete cohort failure at the watershed scale.

Fortunately, in part due to the conservation breeding program for Russian River Coho, all populations impacted by the dry 2013-2014 were able to return to their breeding sites in subsequent years. Oceanic populations and diversities in life history among individuals (i.e. what year the fish return to their spawning grounds to breed) meant that there were still breeding adults who could find their way back to the pools in Northern Californian rivers in subsequent, wetter winters.

Changes in phenology are known to be a challenge for ecology caused by climate change - famously changes in flowering time for plants versus migration or maturation times for pollinators are a concern. This study showed how mismatches between breeding phenology and rivers flows can impact fish species. While Australia doesn't have native salmonids, we do have anadromous fish - like the incredible southwestern lamprey - which also migrate up rivers and may be vulnerable to such changes.