A broader perspective on plant domestication and nutrient and carbon cycling


  • William K. Cornwell,

    Corresponding author
    • Systems Ecology, Department of Ecological Science, Faculty of Earth and Life Sciences, VU University, Amsterdam, the Netherlands
    Search for more papers by this author
  • Johannes H. C. Cornelissen

    1. Systems Ecology, Department of Ecological Science, Faculty of Earth and Life Sciences, VU University, Amsterdam, the Netherlands
    Search for more papers by this author

(Author for correspondence: tel +31 20 5983693; email w.k.cornwell@vu.nl)

A general understanding of plant effects on the global carbon (C) cycle includes both natural and human dominated parts of the world. Furthermore, the full story must also encompass both plant species that evolved without strong human influence and those whose traits have been shaped purposefully by humans – domesticated crops. Dating back to Darwin (1859), scientific progress has often been made by comparing and contrasting the work of natural selection with the process of domestication, as the first two chapters of On the origin of species are titled ‘Variation under domestication’ and ‘Variation under nature’. In another contribution in this long and fruitful tradition, García-Palacios et al., in this issue of New Phytologist (pp. 504–513), have taken concepts and methods developed to study litter quality and decomposition in wild plants and used them to great effect in the study of domestication. Their main result is that litter produced by domesticated species decomposes faster compared to their wild relatives.

‘The first evidence from García-Palacios and colleagues certainly provides some intriguing puzzle pieces.’

A spiral of knowledge: from crops to wild plants and back to crops

Linking wild plant species and their domesticated relatives forms an especially symmetric contribution: the whole field of plant effects on biogeochemistry started as work in agricultural systems (von Liebig, 1843; Richardson, 1938; Silvertown et al., 2006) and then was extended to wild plants by others.

One early milestone in functional plant ecology was ‘The mineral nutrition of wild plants’ by Chapin (1980). Note that the inclusion of ‘wild plants’ in the title was to distinguish this work from that on crop species. The ideas drawn in part from earlier agricultural studies were applied to wild systems where many of the processes are analogous but evolution (not humans) shaped the traits of the plants and competition (not humans) chose the winning species.

One of the more important recent advances that has emerged from the organized study of ‘wild plants’ was the recognition of the leaf economic spectrum (LES; Reich et al., 1999; Wright et al., 2004). Species have very different leaf physiologies, different morphologies, and different rates of key processes, such as photosynthesis. Successful wild plants – those that succeed and maintain viable populations in nature – empirically only have certain combinations of trait values. Species that have long leaf life-spans also have low rates of maximum photosynthesis, low nitrogen (N) concentrations, and high leaf mass per area (LMA). These species can be thought of as seeking – in leaf economic terms – a safe but slow return on C invested in leaves. At the other end of the spectrum we find fast photosynthesis, high N, and low LMA – fast-return species.

This linked set of traits, crucial for C gain while the plant is alive, also has important ‘afterlife’ implications after the plants, or their leaves, die (Cornelissen et al., 2004). Because resorption is incomplete, high-N leaves tend to produce high-N litter, and many recalcitrant compounds (e.g. lignin) are not resorbed and therefore remain in the litter for decomposers to cope with. The effect of these traits on decomposition is not minor: it is often larger than a 10-fold difference in half-life across species at a given site (Cornwell et al., 2008), a substantial effect on biosphere–atmosphere C balance. The links between the role of the traits while the plant is alive and the after-life effects creates the potential for feedbacks (Chapin et al., 1986) with global-scale implications (Heimann & Reichstein, 2008).

One aspect of the theory built to explain the global LES is that a fast rate of C capture and a long leaf life-span are not simultaneously possible in nature (for theory on this see fig. 7 in Reich et al., 1999). Although it has not been studied extensively, most current evidence points to natural selection and not genetic constraints as driving this pattern (Donovan et al., 2010). One part of the proposed explanation for why selection would act in this direction is that a leaf packed with N-rich photosynthetic enzymes and low in secondary compounds would photosynthesize at a high rate but also be very tasty and nutritious to herbivores (Sterner & Elser, 2002), and unlikely to achieve a long life in the real world (Reich et al., 1999). In this way, herbivores behave much like detritivores, and it follows that herbivores prefer to eat the same species that produce highly decomposable litter. Empirical evidence supports this idea (Grime et al., 1996; Perez-Harguindeguy et al., 2003).

Next steps: combining the applied and the basic

Now García-Palacios et al. have followed the spiral of scientific inquiry from crops to wild plants and back to crops, bringing our attention to the domestication process which we now know produces crops whose litter decomposes even faster than their wild relatives’ litter. This is a very interesting finding which raises both basic and applied questions.

On an applied level, there is the question of the extent of land that is devoted to agriculture and how domestication is affecting C balance. The work here on decomposability is one piece of a very large puzzle with many uncertainties. Cultivation practices are complex and a number of issues are relevant. (1) It is of interest whether leaves have first gone through nutrient resorption (e.g. to supply the seeds), a concept frequently studied in wild plants (Aerts, 1996) but rarely in crops. (2) In wild plants, the decomposabilities of (fine) stems and roots are correlated with the decomposability of leaves (Freschet et al., 2012); whether this correlation is maintained in crop species is unknown. (3) The plant litter may also be left on the surface for example with no-till practices or worked into the soil (Gale & Cambardella, 2000). (4) The proportion of biomass harvested relative to total biomass including roots and stems, is of great interest in many agricultural systems (Hay, 2008). In some agricultural practices where the litter is removed, or burned, the differences in litter decomposability will be irrelevant to the C balance, whereas in others when litter is left to decompose on, or in the soil, the full effect of different decomposabilities are likely realized. All these issues have implications for soil fertility and organic matter accumulation; the latter (especially in dry regions) also being important for soil water retention. Putting all of these pieces into a predictive and tunable model of agricultural C, nutrient and water balance is an important goal, especially given the increasing amount of land devoted to domesticated species.

A basic question raised by García-Palacios et al.'s result is the extent to which domestication process parallels natural selection in terms of the traits that affect C cycling. Evolutionary work on the LES has found selection to maintain the linkage between traits (Donovan et al., 2010), that is, the evolutionary driver of the existence of the spectrum itself. García-Palacios et al.'s result is that crop plants are more decomposable compared to their close wild relatives: a first glance interpretation would be that the domestication process moves species along the LES towards the fast-return end. However, there are two lines of evidence against this interpretation: first, although lignin is lower in the domesticated crops, [N] is not as high in crops compared to their wild relatives as you would expect from a shift along the LES (fig. 1 in García-Palacios et al.); second, the litter [N] for the wild relatives are already quite high on a global scale (Supporting Information table S3 in García-Palacios et al.) suggesting that the wild relatives already started at the fast-return end of the LES, and there may be diminishing returns to increasing [N] further. The reduction in lignin concentration during domestication which is not typically included in the LES (sensu Wright et al., 2004), but does in practice reduce herbivore resistance (Wardle et al., 2002), is intriguing. All told, a full understanding of domestication effects on the traits related to the C cycle awaits, with photosynthetic rate, resistance to herbivores, and leaf life span especially intriguing traits to complete the picture. The first evidence from García-Palacios et al. certainly provides some intriguing puzzle pieces.

As we have pushed plants in various ways to maximize yield and aid ease of cultivation, we have created an environment which is very different from where the success or failure of wild plants takes place. Competition is virtually absent, soil resources are supplied in abundance, and herbivores are minimized, through chemical and other means. At first glance for these predominantly annual species, the domestication process might be thought of as sliding along the LES towards the fast-return end and producing litter that is much more decomposable, but there is evidence that the full story may be more complicated. The key difference between nature and crop systems is that in nature there are always herbivores and there are always plant defenses, many of which are constitutive (Moles et al., 2013) and remain in the litter. Reducing herbivores, supplying nutrients in abundance, and selecting for yield and taste may produce organisms that would never survive in nature. The ecosystem-level effect of the traits of domesticated crops may have some parallels in the natural world, but we should also be prepared to find that ‘survival of the fittest’ in nature and ‘selection for human desires’ may yield asynchronous results for the plant traits that affect C and nutrient cycles.