Measuring and modeling roots, the rhizosphere, and microbial processes belowground

Authors


Organized Session – 96th Annual Meeting, Ecological Society of America in Austin, Texas, USA – 10 August 2011

As the field of ecology has progressed, one would be hard pressed to find an empiricist who has not worked with a model or a modeler who has not spent significant time in the field. Still, clear divisions exist between these communities, and despite efforts within both groups to bridge these divisions we still see a distinct break in communication between the two. Part of this disconnect likely comes from fundamentally different aims, as empiricists often highlight novel detail while modelers necessarily look for broad, consistent patterns. Both angles are appropriate and important, but there must be a common ground where empiricists can relay understanding of ecological processes at appropriate scales and modelers can be open to incorporating these insights into new model applications. The gap between empiricists and modelers is particularly evident for the belowground environment, where many ecosystem and most global terrestrial models have been forced to greatly simplify many critical ecological processes due to a lack of empirical understanding and broad descriptive patterns. However, recent advances in our understanding of belowground processes now enable us to push the boundaries of models at both the ecosystem and global scale. Here, models have the potential to improve our descriptive and predictive capabilities in the long term and, perhaps more importantly, can help us to identify key areas of weakness in our empirical understanding that are ripe for focused research in the short term. In a recent session at the annual meeting of the Ecological Society of America, speakers from both the field-based and modeling communities came together to share cutting-edge belowground research and discuss potential new areas of mutually beneficial research.

Belowground – critical, but complicated

Belowground processes control, in part, terrestrial biogeochemical cycles by affecting nutrient and water availability, which control ecosystem productivity (Chapin et al., 2002). Ecosystem productivity in turn influences functions such as carbon (C) storage and nutrient cycling. Therefore, understanding controls of belowground processes is integral to accurately parameterize ecosystem- and landscape-scale models that may be useful for predicting changes in biogeochemistry both across systems and with changes in climate. However, complex interactions between edaphic factors, microbial communities and plant roots that are found in soils often lead to conflicting results across systems that cloud our understanding. Illustrating some of the challenges belowground, Kirsten Hofmockel (Iowa State University, USA) and David Eissenstat (The Pennsylvania State University, USA) highlighted cases where responses to external factors such as elevated atmospheric CO2 and increasing temperature can be species-specific and that even closely related species may differ across dissimilar biomes.

Due in part to the complex and relatively poorly understood interactions belowground many belowground processes are left out of models (Ostle et al., 2009; Iversen, 2010). Still, recent progress in belowground ecology and modeling has been made. For example, the potential for nitrogen (N) limitation feedbacks was incorporated into the Community Land Model (CLM) which resulted in a substantial decrease in predicted plant responses to elevated atmospheric CO2 (Thornton et al., 2007). Continuing further, Joshua Fisher (Jet Propulsion Laboratory, Los Angeles, CA, USA) discussed modeling the C cost of N uptake in the Fixation and Uptake of Nitrogen (FUN) module. In addition, Fisher discussed the potential for a more detailed description of belowground mechanisms based on a conceptual framework put forth in the session by Richard Phillips (Indiana University, USA). In this Rhizo-Accelerated Mineralization and Priming (RAMP) framework, increased N mineralization associated with elevated atmospheric CO2 can be explained by a priming effect through roots and mycorrhizal fungi leading to increased breakdown of recently fixed C. Also important for N uptake and plant productivity, Colleen Iversen (Oak Ridge National Laboratory, TN, USA) described shifting patterns of root allocation in a temperate deciduous forest in response to elevated CO2. These results helped to explain the whole-forest response to elevated CO2 and may provide important details for modeling forest response and soil C storage with climate change.

Challenges continue to persist as large uncertainties surround much of our belowground research. While Erik Hobbie (University of New Hampshire, USA) presented evidence that ectomycorrhizal fungi use organic N of varying age in the soil, he also emphasized that the total amount used by these fungi was still unclear and that the proportion would vary greatly (though not necessarily predictably) with different fungal taxa. Similarly, John Hobbie (Woods Hole Marine Biological Laboratory, MA, USA) provided evidence that we should reconsider how we quantify bacterial activity, arguing that measured amounts of amino acids may be artifacts of sampling methods and that these communities are likely C-starved and yet persist in a state of inactive ‘excess biomass’ (Kuzyakov et al., 2009; Hobbie & Hobbie, 2011). Furthermore, while our understanding of plant roots and identification of root trait patterns has advanced tremendously in recent years (Guo et al., 2008; Xia et al., 2010; Holdaway et al., 2011), Dali Guo (Peking University, China) pointed out that many of our studies do not accurately compare roots of the same type or class and that future breakthroughs will be contingent up the establishment of consistent sampling methodology and larger, cross-species and cross-biome datasets.

Taken together, it quickly becomes clear that simply plugging existing data into models will not be useful without a careful filter through which broad understanding can be gleaned and further research into areas where consistent patterns are lacking. Arguably, modeling of belowground complexities may best be accomplished using different approaches than those that have been commonly used previously. Ray Dybzinski (Princeton University, NJ, USA) proposed that because of the frequency of species interactions occurring belowground, models should take a game theory approach. However, models still need to be grounded both in core ecological principles observed in the field and in mechanistic, or unavoidable, tradeoffs and not apparent tradeoffs that are merely correlative.

Looking forward

There is reason to expect that we will be able to untangle some of the complexities belowground in the near future and inform models more accurately. Widespread availability and adoption of many core belowground methods including genetic and molecular (McGuire et al., 2010; Burke et al., 2011), direct observation of the soil environment (using rhizotrons and root windows; Pritchard et al., 2008; Vargas & Allen, 2008), and measured patterns of stable and radioisotopes (Richter et al., 1999; Lilleskov et al., 2002; Paterson et al., 2009) are building a broad foundation of data that can be used to identify patterns and mechanisms of belowground activity. Furthermore, synthetic activities are piecing broadly related studies together to further our detailed understanding of whole systems (Drake et al., 2011).

We are likely at a point where a tremendous amount of information regarding belowground ecology is available for use in developing and improving models at multiple scales. Arguably, the greatest limitation to improving these models is how effectively we can communicate this information across disciplines. Here, our most efficient route towards improving model descriptions and subsequent understanding of belowground ecology is through direct communication (Fig. 1). While this idea is not new, it is certainly important and appears to be gaining traction among modelers and empiricists alike. Recent programs including the first Integrated Network for Terrestrial Ecosystem Research on Feedbacks to the Atmosphere and ClimatE (INTERFACE) workshop (Leuzinger & Thomas, 2011) and several projects from the National Center for Ecological Analysis and Synthesis (NCEAS) have helped to bring these communities together for information exchange. These efforts should pay dividends as improved models will help to identify knowledge gaps and further direct future empirical research.

Figure 1.

Simple pathway of information flow from empirical data collection to model development, model application, and subsequent re-investment in empirical data collection based on weaknesses identified by the model.

Taking off the rose-colored glasses, we acknowledge that making major advances in our understanding of belowground ecology will not be easy. Working towards this end will require tremendous effort with cutting-edge modeling approaches as well as novel field experiments applied at greater spatial and temporal scales. However, sufficient knowledge and data exist already to enable substantial steps forward with proportionally less work if we can effectively communicate the appropriate information between interested users.

Ancillary