Forest C and N cycling

Tree mycorrhizal type is a predictive plant functional trait – a data synthetic approach

I conducted an extensive global data synthesis showing that AM leaf litter decomposes more quickly than ECM leaf litter across temperate forests, yet this pattern does not hold in sub/tropical forests (Keller and Phillips, 2019). I also collaborated on a synthesis of species-specific foliar nutrient resorption patterns, unveiling a similar mycorrhizal type ×biome interaction with respect to foliar nutrient strategies (Zhang et al. 2018). Specifically, we found ECM tree resorbed more N in temperate forests, while AM tree resorbed more N in sub/tropical forests. Together, these results broadly support the MANE hypothesis that mycorrhizal association is an integrative plant functional trait predictive of both plant trait and soil property differences that drive variation in fundamental ecosystem processes. This work laid the foundation for my greenhouse and field studies.

Relationship between plant belowground carbon allocation and nitrogen uptake -- a greenhouse study

Given our findings that ECM leaf litter turns over more slowly in temperate forests compared to AM species, and given ECM species typically grow in more N-impoverished soils, I hypothesized ECM species may allocate more C belowground compared to AM species while across all species there would be a positive relationship between belowground C allocation and plant N uptake. I grew tree saplings of nine species in soils enriched in isotopically-labeled-nitrogen (N) and after several months, pulse-labeled trees with 13-CO2. This approach allowed me to track the rapid allocation of 13-C from foliage to absorptive root tissue, and to examine how belowground 13-C allocation related to plant 15-N uptake as a measure of each species’ N return on C investment. I found large interspecific differences (but no mycorrhizal type differences) in both the amount of C allocated belowground and plant N uptake. However, mycorrhizal type influenced the relationship between belowground C allocation and N uptake, which was positive in AM species (r = 0.42; P =0.001) and negative in ECM species (r = -0.44; P =0.003), suggesting that the relationship between C allocation and plant nutrition is more complex than theory predicts and that efforts to model these dynamics should consider the traits and tradeoffs that underlie these dynamics.       

Root-derived inputs are major contributors to soil carbon in temperate forests

While we didn't find support in the greenhouse for mycorrhizal type differences in the rapid allocation of C belowground, previous research suggests total belowground C inputs may vary between mycorrhizal type in the forest. How? On the one hand, theory suggests distinct plant and fungal traits between AM and ECM species could lead to greater belowground C inputs by ECM compared AM trees due to 1) more extensive, C-rich external mycelial networks formed by ECM fungi compared AM fungal intracellular arbuscules and vesicles, 2) the greater extracellular oxidative enzymatic capacity of ECM fungi, which occurs a significant C cost and 3) evidence that ECM plants have more slowly-decaying litters (Keller and Phillips, 2019; See et al. 2019) and grow in lower nutrient environments (Phillips et al. 2013), thereby increasing the need to allocate C belowground to enhance nutrient uptake. On the other hand, recent work within select temperate forests shows that plots dominated by AM species store more C in their soils, especially at depth, compared to ECM-dominated plots (Craig et al. 2018), suggesting there may be greater belowground C inputs in AM compared to ECM systems. To test these competing hypotheses, we used a dual-isotopic ingrowth core technique to quantify root-derived C inputs into the soil across six temperate ForestGEO sites in plots varying in AM and ECM tree dominance. We found that that root-derived C was 54% greater in AM versus ECM-dominated plots. This resulted in nearly twice as much root-derived C in putative slow-cycling mineral-associated pools in AM compared to ECM plots. Given that our estimates of root-derived inputs were often equal to or greater than leaf litter inputs, our results suggest that variation in root-derived soil C accumulation due to tree mycorrhizal dominance may be a key control in soil C dynamics in forests.  

Coordination between above- and below-ground nutrient strategies drive soil carbon and nitrogen cycling

I fell in love with ecosystem ecology as a framework for thinking about how various components of nature interact with one another, and I've spent a lot of time thinking about how the 'hidden half' belowground relates to better-understood processes and traits aboveground. To that end, I used a multi-site field study to test the hypothesis that trees have coordinated aboveground and belowground nutrient use strategies. Specifically, I tested if conversation of nitrogen aboveground (via foliar resorption) relates to a tree's need to allocate C belowground to acquire soil nitrogen. In line with my work on belowground C allocation, I expected that the C cost of acquiring soil nutrients both reflects and determines the efficiency of nutrient resorption from senescing leaves. On one hand (what might be considered the 'tree-perspective'), high foliar resorption efficiency (NRE) may reduce C investment belowground, resulting in a negative relationship between aboveground NRE and below ground root production. Alternatively, given that NRE dictates leaf litter N concentrations, high NRE reduces litter N return to the soil and should drive greater tree C investment belowground (i.e. the 'soil perspective'). We found evidence of the 'tree-perspective' across two natural forest sites and support for the 'soil-perspective' at a young common garden site. These results suggest that trait coordination can arise from both a 'tree' and 'soil' perspective, owing to differences in energetic trade-offs and the cost of N acquisition from soil.    


Effects of root litter decomposition on belowground nitrogen cycling

Our understanding of how litter decomposition varies by mycorrhizal type and influences soil nutrient cycling is almost entirely guided by leaf litter studies, and yet both theory and data suggest leaf and root decomposition patterns are distinct as a result of differences in both litter properties and the decomposition microenvironment. I deployed a species × soil type field root decomposition experiment to track nitrogen (N) released from 15-N labeled decomposing roots into the soil, microbial biomass and neighboring plants. I also conducted a complementary greenhouse root decomposition study to isolate the effects of species differences in root chemical and morphological traits on root litter decomposition rates. Data analysis is underway – so stay tuned for the results! 


In a cross-biome meta-analysis, we showed AM leaf litters decay faster than ECM litters in temperate -- but not tropical -- forests (Figure 2; Keller and Phillips, 2019)

AM saplings showed a positive relationship between belowground C investment and plant N uptake while ECM saplings showed a negative relationship (Figure 3; Keller and Phillips, 2019)


Soil ingrowth core used to measure root production across six forests in the eastern U.S.


Root-derived C accumulation was greater in AM-dominated plots vs. ECM-dominated plots across 6 temperate forest sites (Keller et al. 2021)

Above- and belowground C and N trait coordination supported the 'tree-perspective' in two (of three) natural forest sites (a) while evidence of the 'soil-perspective' was detected in a young, common garden site (b) (Keller et al. in review)

Field study tracking root-derived N into soil, microbial and sapling tissues at Moores Creek Research and Teaching Preserve, Indiana