When asked if he would like water in his whisky W.C Fields famously remarked that he didn’t drink water because fish procreate in it (his actual words were somewhat racier). Migratory salmon do so in their millions with a great deal of energy, specifically in the gravel beds of high-energy streams. Before spawning, females lash the stream bed with their tails to create a pit or redd in the gravel, in which they lay their eggs to be fertilised by males. Then she fills-in the redd with more gravel excavated from upstream. Salmon spawning grounds are thus easily recognised as pale patches of freshly overturned gravel on a stream bed that also contain lower amounts of fine sediment and are thereby loosened. As well as discouraging bibulous old men from diluting their liquor, it occurred to Alexander Fremier of Washington State University and other American colleagues that here was a noteworthy example of an active part of the biosphere physically intervening in the rock cycle. Not that it comes even close to what humans have become capable of since the Industrial Revolution, but it might be an object lesson in the fragility of what are otherwise the robust processes of erosion. Moreover, since salmon emerged at some time in the past, their actions might help demonstrate that evolutionary events – speciation, adaptive radiations, mass extinctions etc – play a role in transforming geological processes.
Pacific Sock-eye salmon that die shortly after spawning (credit: Wikipedia)
Fremier and colleagues (Fremier, A.K. et al. 2017. Sex that moves mountains: The influence of spawning fish on river profiles over geologic timescales. Geomorphology online publication; doi.org/10.1016/j.geomorph.2017.09.033) modeled the consequences of salmon spawning habits for the critical stress needed to set grains in motion, theoretically and in a flume tank. Based on a significant reduction of the critical stress, models for the evolution on various river profiles in the vicinity of salmon spawning grounds suggest that river beds may cut deeper at rates up to 30% faster than they would in the absence of salmon. Were salmon to be reduced or extirpated through dam construction or overfishing then sedimentation in channels would increase. In some areas of extensive farming of salmon in offshore pens, escape and colonization of rivers would eventually change sedimentation and erosion patterns. The findings vary from species to species, but salmon may have had a significant effect on generally rugged landscapes following their appearance in local ecosystems.
The terrestrial-marine-terrestrial migratory habits of salmon, including the return of adults to their birth rivers to spawn, are uncommon if not unique. Their forbears must have evolved to this behaviour at some time in the geological past, separately in the case of North Atlantic and North Pacific species. The authors suggest that adaptive radiation of salmon may have been favoured by orogenic events in western North America around 100 Ma ago that created the system of fast flowing rivers that salmon favour. In turn, salmon may have significantly influenced Western Cordillera landscapes of Alaska, Canada and the conterminous Unites States. A nice example of the inseparability of cause and effect on the scale of the Earth System.
As the Yalung-Tsangpo River on the northern flank of the Himalaya approaches a bend the rotates its flow by almost 180 degrees to become the Brahmaputra it enters one of the world’s largest canyons. Over the 200 km length of the Tsangpo Gorge the river descends two kilometres between peaks that tower 7 km above sea level. Since the area is rising tectonically and as a result of the unloading that attends erosion, for the Tsangpo to have maintained its eastward flow it has been suggested that an average erosion rate of 3 to 5 km per million years was maintained continuously over the last 3 to 5 Ma. However, new information from the sediments downstream of the gorge suggests that much of the gorge’s depth was cut during a series of sudden episodes (Lang, K.A. et al. 2013. Erosion of the Tsangpo Gorge by megafloods, Eastern Himalaya. Geology, v. 41, p. 1003-1006).
The Yarlung Tsangpo River watershed which drains the north slope of the Himalayas. (credit: Wikipedia)
It has become clear from a series of mountainside terraces that during the Pleistocene glaciers and debris from them often blocked the narrow valleys through which the river flowed along the northern flank of the Himalaya. Each blockage would have impounded enormous lakes upstream of the Tsangpo Gorge, containing up to 800 km3 of water. Failure of the natural dams would have unleashed equally spectacular floods. The researchers from the University of Washington in Seattle examined the valley downstream of the gorge, to find unconsolidated sediments as much as 150 m above the present channel. They have similar grain size distributions to flood deposits laid down some 30 m above the channel by a flood unleashed in 2000 by the failure of a temporary dam caused by a landslide. The difference is that the higher level deposits are densely vegetated and have well-developed soils: they are almost certainly relics of far larger floods in the distant past from the lakes betrayed by the terraces above the Tsangpo Gorge.
By measuring the age of zircons found in the megaflood deposits using the U/Pb methods the team have been able to show that the sediments were derived mainly from 500 Ma crystalline basement in the Tsangpo Gorge itself rather than from the younger terranes in Tibet. There are four such deposits at separate elevations above the modern river below the gorge. Like the 2000 AD flood deposit, each terrace is capped by landslide debris suggesting that flooding and associated erosion destabilised the steep slopes so characteristic of the region. Because the valleys are so narrow (<200 m at the bottom), each flood would have been extremely deep, flows being of the order of a million cubic metres per second. The huge power would have been capable of moving blocks up to 18 m across with 1 m boulders being carried in suspension. It has been estimated that each of the floods would have been capable of removing material that would otherwise have taken up to 4000 years to erode at present rates of flow.
Grand Canyon in Winter (credit: Wikipedia)
Among the best known and certainly the most visited topographic feature on the planet, the Grand Canyon resulted from erosion by the Colorado River keeping pace with uplift of the south-central United States. It is the archetype for what is known as antecedent drainage. Since that uplift is still going on, albeit slowly, the Grand Canyon has been assumed to be a relative young landform. By dating the first appearance of debris from the eastern end of the canyon in sediments at its western limit geomorphologists estimated that incision began around 6 Ma ago. Yet a range of other observations present puzzling contradictions. One means of settling the issue is to somehow to date the uplift radiometrically.
A long-used technique is to determine ‘cooling ages’ of crustal rocks exposed by uplift and erosion, exploiting the way in which rock temperature determines whether or not products of radioactive decay cab be preserved intact. One method uses the tracks of defects produced by electrons or helium nuclei from radioactive decay as they pass through various minerals that incorporate high amounts of elements such as uranium. Above a certain temperature the fission tracks anneal and disappear quickly, while below it they accumulate over time. Quantifying that build-up allows the date of cooling below the threshold temperature to be estimated. Similarly, gases produced by radioactive decay of some radioactive isotopes, such as argon from the decay of 40K or helium from uranium and thorium isotopes, can only stay in their host mineral if it remains cooler than a narrow range of temperatures. As rock rises towards the Earth’s surface, it starts out hot at depth but cools by conduction as it get closer to the surface. For the 1.8 km of uplift of the Grand Canyon and the relatively cool nature of the underlying crust, neither the fission-track nor the 40Ar/39Ar cooling-age methods give meaningful results. However, minerals lose helium at temperatures above about 70°C, so a method based on helium accumulation from uranium and thorium isotope decay is a possible means of assessing uplift timing. But there have been plenty of snags to overcome to make this approach reliable. In the case of the Grand Canyon analytical quality and careful sample collection has given a credible result (Flowers, R.M. & Farley, K.A. 2012. Apatite 4He/3He and (U-Th)He evidence for an ancient Grand Canyon. Science , doi 10.1126/science.1229390)
Flowers and Farley from the University of Colorado at Boulder and the California Institute of Technology, Pasadena, respectively, produced a result that completely overturns previous conceptions. The western end of the Canyon had been incised to within a few hundred metres of modern depths by 70 Ma ago; more than ten times earlier than previously thought. The eastern end has a more complex history that reveals cooling events in the Neogene as well as an end-Cretaceous initiation of uplift and erosion. Their data are consistent with early incision of the Grand Canyon by a Cretaceous river flowing eastward from the Western Cordillera, with a reversal of flow in the late-Tertiary as uplift of the Colorado Plateau began and western mountains subsided. Whether or not this fits with Cretaceous and later geological history of the SW US, is beyond my ken, but you can bet there will be a storm of comment from US geomorphologists once the paper appears in the print issue of Science.