Location and Study System
This study was conducted on the Tanana River floodplain in the Bonanza Creek Experimental Forest (BCEF), a Long-Term Ecological Research (LTER) site in the northern boreal forest near Fairbanks, Alaska. Details of the climate and ecosystem have been published elsewhere (Andersonet al ., 2009; Ruess et al . 2013). Our study system is the actinorhizal RNS that occurs between thinleaf alder (Alnus tenuifolia ) and bacteria of the genus Frankia . A. tenuifolia is an abundant and ecologically important woody shrub throughout interior Alaskan floodplain ecosystems, in which its life history and ecological impacts have been well studied. On the Tanana River floodplain, A. tenuifolia colonizes alluvial deposits within a few years following deposition, forming very dense stands (~10,000 stems/ha) that provide significant quantities of organic matter and N to early primary successional ecosystems (Hollingsworth et al . 2010; Van Cleve et al ., 1993). In later stages of succession dominated by overstory balsam poplar (Populus balsamifera ), and later, white spruce (Picea glauca ), A. tenuifolia persists in the understory, occasionally proliferating to moderate density in canopy gaps (Hollingsworth et al ., 2010).
The actinobacterial genus Frankia is the microsymbiont involved in actinorhizal RNS. Frankia form effective root nodules with plants from eight families and 25 genera within the eurosid I clade of flowering plants (Swensen & Benson, 2008, Pawlowski & Demchenko, 2012). Phylogenetic studies of Frankia using both single and multi-locus approaches robustly indicate that the genus is divided into four clusters differing in host genus specificity. Cluster 1Frankia form root nodules with alders (Alnus spp.) as well as Morella spp., Casuarina spp., and Comptonia spp. (Swensen & Benson, 2008). Alnus -infective Frankia are phylogenetically and functionally diverse. Within Frankia cluster 1, genotypes have been described that differ in ability to form nodules on specific alder species (Markham, 2008, Vemulapally et al ., 2022a,b), to sporulate within host nodules (Cotin-Galvan et al ., 2016; Markham, 2008; Pozzi et al ., 2015), to subsist on specific carbon sources in soil such as leaf litter (Mirza et al , 2009a), and to support the growth and N-fixation of specific alder species (Markham, 2008; Prat, 1989; Sellstedt et al ., 1986). Alder species appear to exert considerable control over the symbiosis. Surveys of nodules in natural habitats indicate genetic differences in symbioticFrankia among alder species, even when they occur in the same sampling site (Anderson et al ., 2009; Pokharel et al , 2010), and bioassay studies have demonstrated the ability of different species to associate with different Frankia genotypes from the same soil (Lipus & Kennedy, 2011). Recent experiments comparing soil and nodule dwelling Frankia using microcosms in which the concentration of specific Frankia genotypes was controlled demonstrate that alder species can selectively associate withFrankia genotypes disproportionately to their relative occurrence in soil, and that the proportion of selected types differs across some alder species (Ben Tekaya et al ., 2018; Vemulapally et al ., 2022a,b).
In ecosystems of the Tanana floodplain, A. tenuifoliademonstrates both genetic and environmental specificity in its association with Frankia . >95% of the nodules we have collected in several prior studies contain Frankia belonging to a single clade (the ‘AT clade’) that clusters within the largerAlnus -infective group but is divergent from other Frankiawithin it (Anderson et al . 2009, 2013; Ruess et al . 2013). This specificity appears to be a function of the host plant: in sites that contained a second species of Alnus , A. viridis , both host species harbored phylogenetically distinct Frankia despite occurring in close proximity to one another (Anderson et al . 2009). Within the A. tenuifolia -associated clade, individual genotypes are distributed non-randomly across successional environments: nodules in early-succession monospecific alder stands almost exclusively contain a single genotype (‘RF7’) not found in later succession ecosystems; nodules from mid-succession forests dominated by balsam poplar mostly contain a nearly even mix of two other genotypes (‘RF1’ and ‘RF3’); and nodules from late-succession white spruce forests contain a decreased proportion of the two mid-succession types, as well as a higher richness and evenness of other genotypes (Anderson et al . 2009, 2013; Ruess et al . 2013). The association of specific genotypes with successional stages is robust: we have observed the same pattern in replicated, spatially intermixed sites representing each stage, and in repeated samples taken several years apart in a given site.
Sites in the present study include one early succession alder site (FPE3), one late succession site dominated by white spruce (FPL2) examined in Anderson et al. (2009; 2013), and one mid-succession balsam poplar-dominated site (BP1) described in Ruess et al.(2013). The early and late succession sites were selected to differ maximally in composition of Frankia in alder nodules, based on prior observations, in order to maximize our chances of detecting differences in soil assemblages. The mid-succession site is part of a long-term field fertilization study which includes plots fertilized with N (100 kg N*ha-1*yr-1 as NH4NO3, ‘+N’) or P (80 kg P*ha-1*yr-1 as P2O5, ‘+P’) for ≥5 years, and unfertilized control (‘CTL’) plots (details in Ruess et al. , 2013). Soil samples were collected from early and late succession sites in July 2008, and from mid-succession sites in August 2009.