Aim: Ammonia-oxidizing archaea (AOA) and bacteria (AOB) are the primary agents for nitrification, converting ammonia (NH4+) into nitrate (NO3-) and modulating plant nitrogen (N) utilization and terrestrial N retention. However, there is still lack of a unifying framework describing the patterns of global AOA and AOB distribution. In particular, biotic interactions are rarely integrated into any of the conceptual models.
Location: World-wide.
Time period: 2005-2016.
Major taxa studied: Ammonia-oxidizing archaea and ammonia-oxidizing bacteria. 
Methods: A meta-analysis and synthesis was conducted to obtain a general picture of global AOA and AOB distribution and identify the primary driving factors. A microcosm experiment was then conducted to assess effects of relative carbon to nitrogen availability for heterotrophic microbes on AOA and AOB in two distinct soils. A mesocosm experiment was further carried out to characterize the effects of plant roots and their arbuscular mycorrhizal fungi (AMF) on AOA and AOB abundances using hyphae- or root-ingrowth techniques. 
Results: Our meta-analysis showed that soil carbon to nitrogen (C/N) ratios explained the most variance in AOA and AOB abundances, although soil pH had a significant effect. Experimental results demonstrated that high cellulose and mineral N inputs increased total microbial biomass and microbial activities, but inhibited AOA and AOB, suggesting microbial inhibition of AOA and AOB. Also, AMF and roots suppressed AOA and AOB, respectively. 
Main conclusions: Our study provided convincing evidence illustrating that relative carbon to nitrogen availability can dominate the abundances of AOA and AOB. Our experimental results further validated that biotic competitions among plants, heterotrophic microbes and ammonia oxidizers for substrate N predominantly control AOA and AOB abundances. Together, these findings provide new insights into the role of abiotic and biotic factors in modulating terrestrial AOA and AOB abundances and their potential applications for management of nitrification in an increasing reactive N world.

Experiment 1. A microcosm study examining the effects of different ratios of carbon and nitrogen inputs on AOA and AOB.

This experiment was designed to determine the effect of resource C/N ratios on AOA and AOB through quantifying the responses of AOA and AOB to differences in the relative availability of organic C to mineral N. Soil samples were collected from two distinct soils: a pine plantation soil (PINE) in Nanjing 32°03' N, 118°46'E, Jiangsu Province, and a vegetable field soil (VF) in Nantong (32° 01'N, 120°51'E), Jiangsu Province. The PINE and VF soils contained 0.9 and 1.1 g.kg-1 of total N, 37.8 and 13.9 g.kg-1 of total organic C, and had pH values of 4.08 and 7.50, respectively. Field soil samples were sieved through a 2 mm sieve before being used for the incubation experiment.

Each soil was amended with organic C (cellulose) and inorganic N [(NH4)2SO4] at four C/N ratios as follows: 10, 25, 50, and 100. For each soil, two levels of N inputs were designed (Low N at 75 and high N at 200 mg N kg-1 soil) and four levels of cellulose were added at each N level (C/N ratios at 10, 25, 50 and 100, respectively). These resulted in 16 treatment combinations with 18 replicates per treatment (2 soils × 2 N levels × 4 C/N ratios ×18 replicates = 288 microcosm jars), plus two controls (that is, soils with no C and no N amendments: 2 soils ×18 replicates = 36 microcosms).

The soils were pre-incubated for one week and the treatments were then applied. Cellulose (powder) were weighed and well mixed into 50.0 g soil (dry equivalent) and placed into a 250 mL jar with a surface area of 22 cm2. (NH4)2SO4 was applied in a liquid form. The soil moisture was adjusted to 65% of the water holding capacity for each soil. All the microcosm jars were randomly placed inside the room with a room temperature (air-conditioned) at 22-25°C. The soil moisture was maintained by periodically weighing the microcosm jars and adding (spraying) distilled water with a syringe to compensate for any weight loss (every 3 days). The soil samples were incubated for about three months. To ensure sufficient O2, the cover for each jar was opened for 10 min each day for the 1st two months and every 4 days in the 3rd month. The jar opening scheme with different frequencies at different times was designed to ensure sufficient O2 for AOA and AOB, because high soil respiration rate at the early stages consumed more O2 (as shown in Supporting Information Figure S7). A randomly selected subset of treatments (three replicates) was used for monitoring microbial respiration.

Sub-samples of soils (each jar representing one sub-sample) were destructively taken at 3, 7, 14, 21, 35, 47, 60, 75, and 90 days after the application of the treatments to determine the amoA gene copies of the AOA and AOB populations, soil microbial biomass C (MBC), and extractable N (NH4+ and NO3-). Soil samples were taken at 7, 35, 60 and 90 days after the application of the treatments and soil DNA were extracted to determine the abundances of bacteria and fungi (more details below). Soil MBC was measured using the chloroform-fumigation extraction method (Vance, Brookes, & Jenkinson, 1987). NH4+ and NO3- concentrations were extracted with 2 M KCl and detected on a flow injection analyzer (Tecator Inc., Sweden). Soil respiration was measured by an incubation-alkaline absorption method (Hu & van Bruggen, 1997).

Fresh soil (0.30 g) was extracted for DNA using MoBio Power soil TMDNA isolation kits (San Diego, CA) according to the manufacturer’s instructions. Real-time quantitative PCR was performed to determine copy numbers of amoA gene of AOA and AOB, and to quantify bacterial 16S ribosomal DNA and fungal 18S ribosomal DNA in the total DNA of the soil sample using iCycler iQ 5 thermocycler (BioRad Laboratories, Hercules, CA, USA). Primer sets amoA-1F/amoA-2R (Rotthauwe, Witzel, & Liesack, 1997), Arch-amoAF/Arch-amoAR (Francis, Roberts, Beman, Santoro, & Oakley, 2005), Eub338/Eub518 (Fierer, Jackson, Vilgalys, & Jackson, 2005) and ITS1f /5.8s (Fierer et al., 2005) were used for the amplification of bacterial amoA gene, archaeal amoA gene, bacterial 16S ribosomal DNA, and fungal 18S ribosomal DNA fragments, respectively. Real time PCR was performed using the temperature profiles describes in Supporting Information Table S2. Standard curves for real time PCR assays was made as described in Di et al. (2009) for AOA and AOB and Fierer et al. (2005) for bacteria and fungi, respectively.

2. A mesocosm experiment assessing the impact of plant roots and arbuscular mycorrhizal fungi (AMF) on AOA and AOB

This experiment was conducted in the greenhouse at North Carolina State University (NCSU), Raleigh, North Carolina, USA. Two sources of soils [organically managed soil (OM) and conventionally-managed soil (conventional)] were employed to examine how plant roots and their associated AMF affected the abundances of AOA and AOB. These soils were collected from two farming systems at the Center for Environmental Farming Systems at NC State University (35°22'N, 78°02'W) that was established in 1999 in Goldsboro, North Carolina, USA. While the conventional fields had been applied with mineral N, the organically-managed fields had received organic manures and cover crops only since 1999 (Mueller et al., 2006; Tu, Louws, et al., 2006). Both fields were planted with corn (Zea mays L.) prior to soil collection in 2014. Soil samples were partially air-dried and sieved through a 4 mm sieve. The OM and conventional soils contained 24 and 21 μg.g-1 of inorganic N, 178 and 42 μg.g-1 of labile C, and had pH values of 6.50 and 5.50, respectively.

We employed hyphae- or root-ingrowth techniques to examine how the presence of AMF only or roots with their associated AMF affects AOA and AOB abundances in the two soils described above. Plexi-glass mesocosms were used to manipulate roots and/or mycorrhizae, and each microcosm was divided into six compartments with each compartment measuring 13×14×15cm (width × depth × height) (Tu, Booker, et al., 2006). Three compartments in a row were designated as HOST compartments (containing host plants inoculated with AM fungi) and the three adjacent compartments were designated TEST compartments to assess the effects of AMF and/or roots on AOA and AOB. The HOST and TEST compartments were separated by a replaceable mesh fabric panel (Tetko/Sefar mesh, Sefar America, NY) that prevented plant roots or both roots and AMF hyphae from growing into the TEST compartments, respectively. Consequently, this leads to three treatments of AMF and/or roots based on whether the TEST soil is accessible by roots and/AMF from plants in the HOST compartment: (1) the control with no penetration of plant roots and AMF to the TEST soil (CK), (2) penetration of AMF hyphae to the TEST soil (AMF), and (3) penetration of both AMF hyphae and plant roots to the TEST soil (Root). Three different mesh screens, that is, 0.45 µm, 20 µm, and 1.6mm, were used for the Control, the AMF treatment, and the Root treatment, respectively.

The AMF inoculum was a mixture of multiple AM fungal species that were trap-cultured from an agricultural soil, collected from the Center for Environmental Farming Systems at NC State University, and was then pot-cultured to increase fungal biomass. Twelve AM fungal species were identified and characterized according to the International Culture Collection of (Vesicular) Arbuscular Mycorrhizal Fungi (INVAM) (Tu, Booker, et al., 2006). The AMF inoculum consisted of culture media containing spores, hyphae, and colonized root pieces.

Each compartment was filled with 3.5 kg soil mixed with 100g AM fungal inoculum. Organically managed soil was applied with chicken manure at a rate of 12.5 t ha-1, while conventional soil was supplied with urea at a rate of 180 kg N ha-1. The chicken manure (Microstart60, Perdue AgriRecycle LLC) was a pellet with 3-2-3 for N, P2O5 and K2O, finely grounded and mixed with soil before the corn seeds were sown. For the conventional soil, half of urea fertilizer (dissolved in deionized water) was applied to the soil prior to seed sowing and the rest half was applied 4 weeks later. Each time, urea was and then applied to the soil. All sides of all mesocosm units were covered in aluminum foil to block the lights throughout the experimental period.

Four corn seeds (Zea mays L., variety F1 Incredible) were sown to each of the HOST compartments on the 4th March 2014 (days after sowing 0 day: DAS0). The mesocosm units were placed in a temperature-controlled glasshouse by a randomized block design. All of these treatments were replicated three times. The mean daily temperature was 28 ± 2 oC. To ensure 16-h daylight for corns, overhead lights (400 W, Son-T Agro) were used during the early evening hours. The photosynthetically active radiation flux was 505 m mol m-2 s-1 with a day length of 16 h. The plants were watered with deionized water as needed.

The corn plants were allowed to grow for 66 days. All plants were then harvested, dried and weighed, and subsamples were analyzed for C and N contents. Soil samples from the TEST compartments were collected for analyses of AOA and AOB. Fresh soil (0.50 g) was extracted for nucleic acid using FastDNA SPIN kit (MP Bio, Solon, OH, USA) according to the manufacturer’s instructions. Quantitative real-time PCR was performed on each soil sample (CFX96 Real-Time PCR Detection System, Bio-Rad, Hercules, CA, USA) to determine the amoA gene copy numbers of AOA and AOB with the primer sets CrenamoA23f and CrenamoA616r (Tourna, Freitag, Nicol, & Prosser, 2008) and amoA-1F/amoA-2R (Rotthauwe et al., 1997), respectively. Standard curves for real-time PCR assays were made following serial dilutions of the plasmid DNA method (Qiu et al., 2018).