Nitrogen Mass Balance in Caves Ecosystems

1. Introduction:  

An ecosystem consists of all the organisms and the abiotic pools with which they interact. Ecosystem processes include cycling of nutrients and flow of energy from one pool to another. Energy enters an ecosystem through photosynthesis (carbon fixation in primary producers through light), or other external inputs of organic matter drives. The energy is lost from ecosystem when organic matter is oxidized and carbon exported to the atmosphere in form of CO2 through plants respiration, animal and microbial respiration. Nutrients are cycled through both biotic and abiotic pools. Nutrients fluxes involving abiotic components include geochemical processes such as rocks weathering, evaporation, and dissolution. Fluxes involving biotic components include plant uptake, microbially driven processes such as denitrification, dead organic matter oxidation, sulfate reduction, methanogenesis… However, abiotic transformations such as rock weathering may be kinetically enhanced by microbes. Caves are unique environments where sunlight energy and nutrients sources are limited.

The mass-balance approach is used in ecosystem ecology to measure changes in inputs and outputs over time. By selecting a specific nutrient or element and follow its transfer throughout a system, ecologists are able to predict any change in nutrient cycling caused by a perturbation (excess nutrient load from agriculture, species extinction, species invasion…). The mass-balance approach is a very powerful tool in measuring the stability and the resilience of a system to external perturbations, but can also be used as a comparative tool to assess differential responses to a given perturbation among various ecosystems; however, the mass-balance approach is limited in explaining mechanisms causing the changes.

1. Description of target ecosystem:

Our target ecosystem is typical epigenic cave aquifers under agricultural landscapes. Caves originate from the dissolution of carbonates (limestone, dolomite, marble) by the action of acidic waters. Caves are huge groundwater reservoirs. More than 25% of the world groundwater reserve comes from cave aquifers. Caves have also played a major role in historic human settlement as major cities were built around springs (e.g. Colorado Springs). Caves are habitats for diverse species (over 1,138) both terrestrial and aquatic that possess the unique ability to live in perpetual darkness and limited food resources conditions (Kendall et al. 2015).

At a global scale, carbonates dissolution acts as a major sink for atmospheric carbon and recent scenario are considering carbonates systems as an alternative C sink to reforestation and ocean storage (J. Cao et al. 2016). Main nutrient studies in caves have focused on C and S interactions due to their link to carbonate dissolution and importance in global C dynamics (Bennett and Engel, 2005; Engel et al., 2004; Engel et al., 2001). Few studies have considered N cycling in caves.

For a long-time, cave systems have been considered as ‘closed systems’ regarding Total Inorganic Nitrogen (TIN) inputs. Many studies have been focusing on cave systems (Jarowshi et al., 1992; McFarlane et al., 1995) with limited TIN inputs from external sources, where Nitrogen is almost essentially provided by bats excretion. Cave aquifers under agriculture and farming land-use represent a unique scenario where N inputs may be dominantly allochthonous. In addition, recent studies in Southern Indiana and South-Central Kentucky caves (Carr, 2011; Toomey and Thomas, 2011) indicated a significant decline in bats population in response to the White-Nose Syndrome (80% decline in little brown bat, above 80% declines in Indiana bats, and near 70% decline in tricolored bats), which may affect N inputs from bats. How will the decrease in bat populations affect N cycling in caves? How do increasing inputs from agricultural land-use affect N cycling within cave aquifers? How much of the N cycling in caves is understood and how much is still to be explored?

2.1. Community structure and Ecosystem Function:

As of ecosystem function, we are focusing on dissolved inorganic nitrogen; more specifically on NO3 fluxes since ammonia (NH4+) fluxes can be neglected in the overall inorganic nitrogen flux within normal stream conditions. Various studies conducted in basins with different superficies, different land-uses and under different climates, reveal [NH4]/[NO3] ratios varying within a narrow range (0.01 to 0.03). In terms of community structure, we will consider two main components: a biotic component, consisting of autotrophic and heterotrophic bacteria, macroinvertebrates and mammals, and abiotic variables such as CO2 pressure, temperature, pH, Dissolved Organic Matter (DOM) concentration, redox conditions.  CO2 pressure (dissolution of atmospheric CO2 in percolating water) combined with temperature and pH will affect CaCO3 dissolution and bacterial metabolism; coupled to redox conditions (anaerobiosis) will affect the rate of denitrification.

2.2. TIN pools:

We defined three main pools of Inorganic Nitrogen:

-The groundwater pool: An average N background concentration value range between 2 and 2.5 mg N-NO3/L has been established throughout the Midwest in carbonate stream water systems (Panno and Kelly, 2003; Panno et al., 2006). An average value of 2.38 mg/L has been determined for the Grayson-Gunnar Cave (GGC) system in SE Kentucky (Tagne and Florea, 2015). Considering an 11km-long cave stream (GGC), and N/NO3 ratio of 0.226, this corresponds to a pool of 5.108 kg N/ha

• The macroinvertebrates pool: Barratt et al. 2009 estimated the macroinvertebrates using the total population of arthropods (4.7*104 mg N/m2 = 4.7 102 kg N/ha)
• The microbial pool is composed of both heterotrophic and autotrophic bacteria.

• Heterotrophs: If we assume that microbes population that are involved in dissimilatory sulfate reduction in cave systems is similar in size to the microbial pool involved in nitrate reduction, then we can estimate an heterotrophic pool of 0.043 mg N/dm3 (Aharon and Fu, 2000), which is equivalent to 3.9 107 kg N/ ha (considering a 11km-long cave stream (GGC)).
• Autotrophs: Assuming that the heterotrophic pool represents 1.7% of the autotrophic productivity (Engel et al. 2004), we can estimate the autotrophic pool at 2.6 109 kg N/ha.

2.3. TIN fluxes:

2.3.1. TIN inputs:

 Inorganic Nitrogen inputs are divided between terrestrial/allochthonous inputs and bats inputs.

• Terrestrial inputs are estimated at 2.76 kg N/ha/yr (see conversion Tab.1). TIN fluxes are obtained by dividing annual inputs (in kg NO3-N; Panno and Kelly, 2003) by the total basin area (3770 ha) and multiplying by the N/NO3 ratio (0.226).
• Bats inputs: N flux through bat excretion ranges between 1.95 and 2.52 mg N/m2/day (Welbourn, 1999). Considering the current average extinction rate of 80% due to the White-Nose Syndrome (Carr, 2011; Toomey and Thomas, 2011), we can deduce actual N inputs from bats at 0.4 mg N/m2/day, thus 1.44 kg N/ha/yr (~10% of terrestrial inputs).

2.3.2. TIN cycling:

- Chemoautotrophic microbial metabolism: The activity of chemoautotrophic sulfur-oxidizing communities in cave systems has been widely addressed in the literature (Bennett and Engel, 2005; Engel et al., 2004; Engel et al., 2001). Little attention has been given to the microbial activity on N transformations within cave aquifers. However, based on models suggested for C and S transformations, we can predict a model of autotrophic denitrification.