Introduction
A key prediction in evolutionary biology is that relatives will cooperate more and compete less (Griffin & West 2002; West 2002; West et al. 2007). Kin selection emerges because relatives share high proportions of their genes, and by cooperating, more of these genes are passed to the next generation. Kin selection has been applied to understand a diversity of cooperative phenomena, from cooperating RNA molecules (Levin & West 2017a, b) to cooperation among human beings (dos Santos & West 2018). However, kin selection can also be vulnerable to competition among relatives, especially in cases where there is high spatial structuring in the population. Under these cases, increased competition among relatives can totally negate benefits of cooperation (Queller 1992; West 2002).
While there is an increasing understanding of when and why relatives cooperate, it is unknown how these dynamics affect organisms interacting with a group of relatives. In symbiotic partnerships, for example, a host interacts with communities of microbes, which can vary in levels of relatedness (Fosteret al. 2017). The host may benefit from interacting with highly related strains because of reduced conflict and competition within the community (Frank 1996a, 2003; West 2002). However, interacting with less related strains may not always entail a cost for the host, and could even be beneficial. Specifically, if there is a greater relative difference among the symbiont species in their ability to acquire different, or complementary resources, the host could benefit from interacting with non-relatives (Jansa et al. 2008; Wagg et al. 2011). Likewise, if competition drives an underbidding scenario, which results in symbionts providing more benefits, for less in return, the host could benefit from interacting simultaneously with competing symbiotic strains (Wyatt et al. 2014; Noë & Kiers 2018).
Manipulating relative relatedness in symbiotic communities has historically been challenging, making direct tests of these ideas difficult. Here, we use the arbuscular mycorrhizal symbiosis to study the effects of symbiont relatedness on host plants interacting via competing or cooperating mycorrhizal fungal networks. The vast majority of land plants are colonized by arbuscular mycorrhizal fungi. The fungi trade soil bound nutrients such as phosphorus and nitrogen for photosynthetic carbon from the host plant (Jiang et al. 2017; Keymer et al. 2017; Luginbuehl et al. 2017). The fungi form underground networks that can connect roots of different plant individuals. Hyphal fusion, otherwise known as anastomosis, can occur among closely-related arbuscular mycorrhizal fungi (Giovannetti et al.2004; Jakobsen 2004). This has the potential to increase resource sharing across the fungal network (Johansen & Jensen 1996; Walder et al. 2012), which could increase the fitness of the fungi (Giovannetti et al. 2015) and potentially their hosts (Roger et al. 2013). However, when fungi are genetically less related, the hyphae can be vegetative incompatible and fusion will not occur (Giovannetti et al.2003; Croll et al. 2009). Direct antagonism among competing arbuscular mycorrhizal strains has been shown to lead to negative outcomes for fungal abundance and plant growth (Engelmoer et al.2014), and can also influence fungal co-existence within host roots (Roger et al. 2013). For example, past work has shown how competition between distantly related arbuscular mycorrhizal fungal isolates resulted in almost complete exclusion of one isolate by the other, whereas more related isolates shared the roots space in an almost 50:50 proportion (Roger et al. 2013). While this suggests that level of relatedness can affect fungal competitive dynamics within a root, it is unknown how relatedness affects the functioning of the hyphal network, especially when the hyphae connect multiple plants.
Our aim was to understand how fungal relatedness affects the physical formation and nutrient transfer in a fungal network formed between host plants. To study phosphorus distribution and transfer, we employed a recently developed technique in which we tag phosphorus rock (apatite) with fluorescent quantum-dot nanoparticles (van ’t Padje et al.in press; Whiteside et al. 2019; van’t Padje et al. 2020). Quantum-dots fluoresce in bright and pure colors when excited with UV-light. We used a class of quantum-dots that were highly fluorescent, stable and well characterized in terms of toxicity, uptake and transfer by fungal hyphae, with accumulation patterns in root (and leaf) tissue as expected (Whitesideet al. 2009; Gustafsson et al. 2015). This allowed us to determine how much phosphorus was transferred across the fungal network per unit of fungal biomass.
We grew a host root colonized by a single focal strain. The arbuscular mycorrhizal hyphae of this focal plant were allowed to interact with a fungal network of a second host plant that was either the same fungal strain (“selfing”) or two genetically less-related fungal strains (both “non-selfing”). In order of highest to lowest relatedness, these treatments included: (i) the same fungal strain (selfing), (ii) a different fungal strain within the same species (non-selfing), or (iii) a fungal strain of a different species in the same genus (non-selfing). We grew these plant and fungal treatments as both whole plants in soil and as in-vitro root organ cultures in petri plates. The latter allowed us to determine where phosphorus was distributed across the network using our quantum-dot tagging technique, as well as to quantify the physical fungal network structure using imaging techniques (Boddy 1999; Heatonet al. 2012a, b). We determined how varying relatedness in fungal networks between the two host plants influenced: (i) host growth, (ii) fungal colonization inside root tissue (intraradical colonization), (iii) network formation outside the root tissue (extraradical colonization), and (iv) transfer of nutrients across the network to the host root.