1. Toby Ahrens 27 Sept 2004 Linking Spatial Variability of Soil N Retention Mechanisms to Landscape-level Fates in Yaqui Valley, Mexico

2. The hook…  Where do I want to be? Quantify degrees of N contamination under different management regimes Investigate the value of spatial data sets varying in resolution Probability of “dangerous” N application  What do I need to get there? Link process and transport models Improve process models  Include abiotic retention mechanisms

3. Yaqui Valley, Sonora, MX

5. Coastal eutrophication Declassified Keyhole satellite image March 8, 1978 (Beman, unpublished data)

6. Nitrate-contaminated groundwater <10 ppm 10-40 ppm 116 ppm

7. Two Q’s guiding my efforts…  What soil characteristics control N availability, retention and loss in OM- deficient ag soils?  Can the spatial variability of retention-dominating characteristics be linked to aqueous N fates throughout the Valley?

8. Watershed-level aquifer vulnerability studies: not mechanistic…  Watershed-level aquifer vulnerability studies use empirically-derived equations to predict leaching… Makes management extension hard!  Process- and index-based models not spatially explicit Makes fate + transport link hard!  RHESSys used similar reasoning for natural systems

9. Model schematic… soil properties NLOSS soil mineralogy crop yield (+N use) Solute transport un/saturated boundary conditions management unit crop type Output maps: Leaching vulnerability Aquifer contamination Coastal N sources groundwater depth

10. NLOSS  Nitrification: hole-in-the-pipe sequence  Denitrifier kinetics: double Monod  Decomp: 3+ pools, not important to denit.  Outputs: Leached N, trace gases  Soil moisture: Solve Richard’s equation Accounts for macropore plow (using Bypass model of Eckersten and Jansson 1991) Mechanistic treatment of anaerobic fraction

11. NLOSS, cont’d  No sorption or fixation NH4+ either taken up by plants, immobilized, or nitrified.  Resolution in unsaturated zone needs to be 10 cm or less

12. Lee’s model…

13. Major modeling efforts:  Soil submodel including sorption isotherms and mineral fixation  Spatial data referencing  Solute transport component to saturated hydrology model  Saturated/unsaturated boundary layer conditions

14.  What soil characteristics control N availability, retention and loss in OM- deficient ag soils?  Can the spatial variability of retention-dominating characteristics be linked to aqueous N fates throughout the Valley? Two Q’s guiding my efforts…

15. Applied N: 250 kg/ha Plant uptake: 31% Leached: 2-5% (14-26%) Gaseous losses: NO+N 2 O: 2-5% NH 3 : __% N 2 : __% ? ? ? ? Drainage canals: NO 3 - + NH 4 + : 2-5% NO+N 2 O: <0.1% References: Riley et al. 2001 Harrison 2003 Matson et al. 1998 Ortiz-Monasterio, pers. comm.

16. Adsorption Reactive N AbioticBiotic Fixation AEC CEC OM Clay minerals Q1: A series of field and lab experiments to identify the major pools, turnover rates and bioavailability of N…

17. Adsorption Reactive N AbioticBiotic Fixation AEC CEC OM Clay minerals Hypotheses… Q1: How much N is immobilized by abiotic vs. biotic processes during an irrigation event? H1: Abiotic mechanisms are significant sinks of ammonium due to moderate CECs, mineralogy, timing of N addition and low availability of labile C for heterotrophs… Study 1: Sterilize with HgCl 2 Add 15 N-(NH 4 ) 2 SO 4 Extract with KCl Dry and grind soils for 15 N analysis Methods: Johnson et al. 2000 Wolf et al. 1999 Preliminary data: Panek et al. 2000 Riley et al. 2001 Heterotrophs Nitrifiers Plant uptake Fungi/myc. Denitrifiers DNRA

18. Adsorption Reactive N AbioticBiotic Fixation AEC CEC OM Clay minerals Hypotheses… Methods: Philips 1999 Methods of Soil Analysis 1996 Preliminary data: Limon-Ortega et al. 2001 Riley et al. 2001 Study 2: CEC: expose soils to range of [NH 4 + ], centrifuge, filter and ion chromotography AEC: same, but use NO 3 - Kinetics: establish isotherms (also extract initial ions…) Q2: How much adsorbed N is associated is exchangeable? H2: The high pHs and presence of 2:1 Si-O clays results in moderate CEC and low AEC… CEC kinetics follow a Langmuir isotherm model… organic mineral

19. Adsorption Reactive N AbioticBiotic Fixation AEC CEC OM Clay minerals Hypotheses… Q3: What is the potential for soils to fix NH 4 +, and is there field evidence of fixation? H3: Soils are rich in montmorillinitic clays, and significant amounts of NH 4 + -N is associated with those minerals… Low OM concentrations leads to low OM-associated fixation Study 3: Mineralogy XRD: particle size sep, then dry slurry on glass slide… Spectroscopy in lab Fixed N Sequential dissolution: size sep then KCl, H 2 O 2, HF OK for labeled/sterile soil… Preliminary data: Osher, pers. comm. Enriquez et al. 1998 Cade-Menun, unpub. data Methods: USDA 1996 Paramaasivam and Breitenbeck 2000

20. NH 4 + fixation in clay minerals Fixed and exchangeable NH 4 + in an illite (Stevenson et al., 1982)

21. Adsorption Reactive N AbioticBiotic Fixation AEC CEC OM Clay minerals Hypotheses… OM fraction Q4: Is there a potential for abiotic reactions between NO 2 - or NH x and SOM? H4: Chemodenitrification and NO 2 - rxns with OM not important due to pH. Don’t yet have an educated guess as to the OM fate of NH 4 +, but I don’t expect much to end up in limited OM… Study 4: OM N pools: steam distillation with acids and bases, recover NH 3 OK for labeled/sterile soil… Methods: Davidson et al. 2003 Stevenson 1982,1996 Preliminary data: Riley et al. 2001

22. Aforementioned studies not exclusive…  Largest pathways will be followed with more detailed studies… Clay mineral fixation: bioavailability or remineralization studies Biotic immobilization: heterotrophic vs nitrifiers, etc. OM fixation: NMR or mass spec to look at actual compounds involved in fixation

23. Future directions and broader implications…  Does better spatial resolution increase our ability to predict SW/GW vulnerability? And the degree of how mechanistic submodels are?  What is the probability of “dangerous” N application? Couple to economic optimization model following the Mastrandrea and Schneider (2004) method…

24. less weathered more weathered

25. Resolution of input data: Field based…

27. Resolution of input data: or NPP driven by remote sensing…

28. Timeline ’04 ’05 ’06 ‘07 Fall Win Spr Sum Fall Win Spr Sum Fall Win Spr Sum Fall Planning, cont’d XXX Field sampling XXX N pools + rates XXX Total C,N, etc. XXX Abiotic vs. biotic XXX CEC, AEC XX NH4 fixation XXX OM fixation XXX Modeling – submodel dev’t XXXXXXXXXX Multiple constraints – 15N? XXXXXXX 15N analyses XXXX Model completion XXXXXXXXXXXXXXX Writing Plant vs. leached N XXXXX Spatial leaching vulnerability XXXXX Value of spatial process modelsXXXXX Dangerous N application XXXXX Graduate X

29. Ivan’s 1 st question: Triticales Durum Pasta

30. Riley et al. 2001…

31. Tractor-mounted hollow stem augers for subsoil sampling http://www.geology.sdsu.edu/classes/geol552/hollowstem.htm

32. Aquifer depth Addams, 2004

33. Clay content in shallow subsurface Figure ‎4 ‑ 28 Estimated fraction of clay in Layer 1 (“Shallow Horizon”). Estimated values for the cells of the model grid range between 0-1 and were interpolated from measurements at well locations (blue). Addams, 2004

34. Anticipated publications  Controls on N availability to plants and N loss to surface and groundwater differ in an agricultural soil in Northwestern Mexico.  Spatial variability of montmorillonite and management history explain contaminated groundwater distribution in an agricultural landscape in Northwestern Mexico.  How mechanistic do process-based spatial models need to be to predict N fates at the landscape scale?  What is the probability of “ dangerous ” N application in an irrigated wheat system in Northern Mexico?

35. Applied N: Plant uptake: Leached: Drainage canals: References: Panek et al. 2001 δ 15 N

36. N oxidation states Davidson, 1991

37. Chapin et al, 2002

38. Major subsurface flowpaths Haag and Kaupenjohann, 2001

39. Corridors and retention zones Haag and Kaupenjohann, 2001

40. Lytropic series  Al 3+ >H + >Ca 2+ >Mg 2+ >NH 4 + >K + >Na +  PO 4 3- >SO 4 2- >Cl - >NO 3 -

41. 26% of area not in wheat production

42. % of area not in wheat  26% in winter growing season (120 day growing season)  ~80+% in non-winter growing season  > 60% of the time, a given area is not is use for wheat production

43. Can what I’ve proposed get me there?  N fates in soil + groundwater: Yes, with flexible approach…  N to estuaries: will need collaboration with Adina’s lab  N modeling Q’s: Yes, keep uncertainties explicit…