Answers are grounded in this paper and cite their sources.
Arbuscular mycorrhizal (AM) symbioses involve transfer of soil nutrients to plants and plant-derived organic compounds to AM fungi. Many experiments have shown that individual symbioses are ‘mutualistic’ (Table 1). However, outcomes in relation to the nonmycorrhizal (NM) state are not always positive for plants, although they are for all AM fungi, which are obligate symbionts. Different AM fungi produce different growth-related responses in individual plants and there are large differences in responsiveness among different plant taxa in relation to colonization by the same AM fungal taxon (Klironomos, 2003). Neutral outcomes for the plant may be termed ‘commensal’, and negative ones ‘parasitic’ (Table 1). Outcomes are influenced by many abiotic and biotic factors (Johnson et al., 1997). Here we address the extent to which AM symbiosis can be validly interpreted as a fundamentally ‘harmonious’ relationship, as is suggested by use of terms such as ‘cooperation’ and ‘rewards’ (Table 1). Although harmony may not be a familiar term, it has been used in relation to plant–microbe symbioses (Broughton et al., 2000; Oldroyd et al., 2005; Martin et al., 2007). The issue is also addressed briefly in a report of the 33rd New Phytologist Symposium, Zurich, May 2014 (Bender et al., 2014). We focus on four concepts in the literature that relate to uptake of phosphate (P), a growth-limiting nutrient in many soils. The concepts are relevant to other nutrients that are taken up via AM fungi, such as inorganic nitrogen (N). We comment specifically on whether increasing use of subjective terminology actually facilitates interpretation of results and will help experimental design in future. This concept considers AM ‘function’ (Table 1) and its diversity in relation to net costs and benefits to plants with respect to the NM state (Johnson et al., 1997). Outcomes are usually measured in pot experiments in terms of plant biomass (mycorrhizal growth response, MGR), or growth-related parameters such as P content per plant, and we previously suggested that considering a continuum of responsiveness would be preferable (Smith & Smith, 2011a). Potential ‘fitness’, for example as measured by plant fecundity and survival (see Table 1), might be a better indicator of position on the continuum as long as experiments include appropriate measurements, which most do not (Johnson et al., 1997; Jones & Smith, 2004; Lekberg & Koide, 2014). Even in pot experiments it is not easy to quantify net costs and benefits to the partners in physiological terms. For example, MGRs do not always relate to amounts of P taken up by the AM fungal pathway. This has been shown by tracking the AM pathway of P uptake and transfer to the plant with radioactive P. The data in Table 2 demonstrate (1) interplay between uptake between the AM and direct (epidermal) root uptake pathways and (2) differences between individual AM fungi. Such results show that zero or negative MGR usually does not result from AM fungal parasitism where AM P transfer would be zero. We suggested that negative MGR may result directly from decreased direct epidermal P uptake that lowers photosynthesis rather than excess carbon (C)-drain to the fungus (Li et al., 2008; Smith et al., 2009). Physiological diversity is also displayed by experiments employing root-organ culture with AM fungi (e.g. Kiers et al., 2011 and references cited therein), though these systems necessarily lack controls exerted from shoots and hence the existence of a natural feedback loop for resource exchange (Fortin et al., 2002; Smith & Smith, 2011b). Physiological diversity is important in helping interpret ecological outcomes where plants are colonized by many AM fungi, especially when AM plants are connected by common mycelial networks (e.g. Walder et al., 2012; Fellbaum et al., 2014). However, there is a major problem in extrapolating the significance of experiments with individual AM plants and fungi to higher scales such as plant communities and ecosystems (Francis & Read, 1995; Lekberg & Koide, 2014). In focusing on growth-related outcomes, the mutualism–parasitism continuum has no subjective elements. It does not even imply cause–effect issues of partner recognition and subsequent resource transfer, though these can be introduced in interpreting the mechanistic basis (Smith & Smith, 2013). Accordingly, Table 1 gives the components zero subjectivity. This situation contrasts to some extent with frequent descriptions of AM resource transfer as resource trade. ‘Trade’ implies an element of cooperation; both terms have low subjectivity (Table 1). By definition, trade does not cover fungal parasitism, where transfer is one-way (C from the plant); mycoheterotrophy is also a complication. Bidirectional resource transfers are increasingly described as ‘rewards’. The term implies higher harmony than does trade, because traders (at least human ones) do not acquire rewards: they seek the most favourable terms of trade. We consider that the rewards concept has high subjectivity (Table 1). ‘Reciprocal’ rewards are claimed to lead to plant-based pressures disadvantaging AM fungi that obtain C but restrict P supply to the host (‘cheaters’: Table 1). So-called ‘preferential rewards’ to ‘high-quality’ partners are said to stabilize cooperation in AM symbiosis, with evolutionary consequences (Kiers et al., 2011). There appears to be an element of circularity, as the rewards (resource transfers) are the mechanistic basis of the cooperation. The AM fungi that transfer the most P or N inevitably require the most C, because uptake of P and transfer to the plant requires energy, while hyphal transfer of N as arginine requires organic C skeletons. In other words, rewards inevitably carry a cost, the significance of which for an AM plant depends on its ability to lower C costs of growth compared with the NM state (Johnson et al., 1997). There is a parallel even in NM plants which extend parts of their root system into nutrient-rich patches, but the C-nutrient trade would not normally be described as ‘reciprocal rewards’ – ‘supply and demand’ would be conventional terms in this context. Cause–effect issues of which reward (P or C) comes first once colonization is established are unresolved. Kiers et al. (2011) commented that AM fungi may ‘choose’ to transfer P to the host that offers the highest C benefit. By contrast, Fitter (2006) suggested that plants do not provide C to fungi that transfer no P. The result should then be inhibition of ongoing colonization if this cheating can be recognized by the plant. Although the degree of colonization can be very variable in pot experiments with individual AM fungi, there is no physiological evidence to our knowledge that low colonization is necessarily linked to a plant response to prevent or minimize fungal cheating. However, physiological measurements suggest that cheating in the extreme sense of little or no P supply in return for C is much less common than previously inferred from lack of positive MGRs (Smith & Smith, 2011b). This may be a consolation for researchers who think of the symbiosis in terms of rewards. It is sometimes suggested that AM fungi that are not the most beneficial in terms of nutrient transfers may confer non-nutritional benefits that result in increased fitness in the presence of stress such as disease. This possibility is feasible over evolutionary time, but it looks contrary to a simple ‘rewards’ concept if a host preferentially rewards a fungus that is not of ‘high quality’ in the absence of the stress in the anticipation that its value increases in the presence of the stress. Resource transfer in AM symbioses can be also regarded as mutual exploitation of the partners’ resources (Egger & Hibbett, 2004), also with high subjectivity (Table 1). Mutual parasitism would be a similar but more extreme term. The exploitation terminology reflects a different mindset than does ‘rewards’ about the physiological aspects of the symbiosis and the outcomes across the mutualism–parasitism continuum of responsiveness, again without necessarily specifying causes and effects. The terminology does not imply AM cooperation or harmony: perhaps that is why it receives little use. Ecologists have used economic theory to consider mutualisms in general as ‘biological markets’ since the 1990s. The concept is being extended to microbes (Werner et al., 2014), including mycorrhizas (Kiers et al., 2011; Fellbaum et al., 2014). The details are summarized in Table 1 and go a long way beyond traditional cost–benefit analyses that have been extensively used in relation to AM symbioses (e.g. Koide & Elliott, 1989; Johnson et al., 1997 and references cited therein; Jones et al., 1998). A problem in applying this concept to AM symbiosis is that there is no monetary price: a bartering process determines ‘rate of exchange’ of resources. ‘Unfair’ trade (cheating) is a complication (Noë & Hammerstein, 1995), as is interplay between direct epidermal uptake via the roots (‘nonmarketing’) and uptake via the AM pathway. The trade is highly context-dependent in relation to the wide range of factors that influence AM physiology and outcomes of individual symbioses along the (nonsubjective) mutualism–parasitism continuum. However, the underlying mindset is less subjective than with ‘mutual rewards’ (Table 1). It is worth pursuing mindsets in more detail beyond issues of harmony. Although the biological market concept by definition focuses on the AM state, much of the terminology used by researchers (including ourselves) uses the NM state as the basis. This applies to MGR: conventionally 100(AM − NM)/NM, in terms of weight or other growth-related parameters such as plant P uptake. An alternative term is ‘dependency’ (or ‘dependence’), which appears more AM-focused, but is again conventionally defined with reference to the NM state: 100(AM − NM)/AM (Habte & Manjunath, 1991). Where dependency is zero (NM = AM) it has been interpreted as implying that there is no AM P uptake but, as emphasized earlier, this is not necessarily the case. In other words, ‘dependency’ must again be defined, and used as an outcome, not a resource-based mechanism. In nature the NM state of a potentially AM plant is usually artificial, unless the soil lacks AM fungal propagules. Perhaps we should really talk of nonmycorrhizal growth response (NMGR). The same NM mindset is reflected in our own use of ‘suppression’ or ‘inhibition’ of direct P uptake when plants receive P via the AM pathway but have zero or negative MGR (Grace et al., 2008; Smith et al., 2009; Smith & Smith, 2011b). More realistically in the (abnormal) NM state these plants can respond by up-regulating direct P uptake beyond that in the AM state. This can be regarded as a response to stress (Smith et al., 2011) and may be highly relevant to the evolutionary success of plants with zero or even negative MGR in low-P soil. The answer to our question in the title is that most AM symbioses can validly be regarded as harmonious in terms of recognition by symbionts and evolutionary success of the symbiosis. Even AM fungal cheating can be regarded as harmonious to the extent that the plant tolerates it (using ‘tolerates’ as yet another subjective term). However, the extent to which there is harmony in competition among AM fungi co-existing in a root, or among plants sharing a mycelial network, is not clear. Accordingly, we are not advocating use of the term. Subjective terminology is inevitable in describing AM symbioses, just as it is with other symbioses. The danger is that it can lead to over-simple generalizations and may then unhelpfully influence subsequent experimental approaches. Our particular concern is uncritical use in interpreting AM physiology in ecological or evolutionary contexts. The market concept is attractive because it deals with partitioning of resources when roots contain several AM fungal taxa and when plants are connected by common mycelial networks. However, it is not clear how much it helps understand AM functional diversity, especially when this extends to mycoheterotrophy (Selosse & Rousset, 2011). The concept may extend impact of research with AM fungi into a wider research arena but a detailed assessment of its value beyond general principles is certainly needed. Irrespective of terminologies, we believe that physiological bases of resource transfer in AM symbioses need to be much better understood in terms of causes and effects, and the range of mineral nutrients that can be transferred. Inevitably, experiments will be reductionist compared with the real world of plant growth in the field, but as long as limitations are recognized in relation to relevance across scales this should not be a problem in moving forward. This letter developed from a poster at the 33rd New Phytologist Symposium (Zurich, May 2014), and we thank the New Phytologist Trust for grants to attend. The title was influenced by the official Symposium cartoon that showed a mycorrhizal fungus and root shaking hands. We showed boxing gloves in the poster, while Bender et al. (2014) illustrated both versions. We are grateful to Nancy Johnson and two anonymous reviewers whose comments were very helpful for revision.
No takes yet. Share an insight, caveat, or question.
SMITH et al. (2014) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: