This file was created by scanning the printed publication.
Errors identified by the software have been corrected;
however, some errors may remain.
• 594·
MLitifU1CtionaJ GrasSands in a Glcqjng World
Voltme I
Procedures for estimation of the livestock ecological footprint in US dlYlands
Ro/)('rt il . W (tshington-A ller/ lLnd Juhn E . Mitchell'
Department of Ecosystem Science antiManagement. Texas.4 &M University.' 138 TilMU ,College Station, Texas 778432138, USA, E-mail: wmhington.aller@tamul.Illl
2 USDA .Forest Service Rock; Mountain Research Station, 2150 Cmtre Ave. , BlIilding A , Fort Collin.., , Colorado 80526 ,
US!1 . E-m<li/ : jemitchelKij). fs led .11."
I
Introduction The ecological condition and trend of the United States':i ~)02 !lOO krn· of Drylaml;; and tlw ex"~nt to which human
management aetions contribute to degradation are unknown at the national spatial seal" (W ashington-AlJen et al . 20(6) . Our
research seeks to develop procedur ... s for determining tlw impact or ecological footprint of livestock grazing on the productive
capacity of Drylands at the national spatial scale.
2
Approach The study approach has been to develop a Geographic Information System (CIS) consisting of digital maps of gross
andne! primary produetivity (CPP and NPP) , lJ .S . Department of ""griculturc National Agric-ultural Statistir.s Office
(NASO) national livestock number and distribution maps. transportation and power c(lIlsumption networks, gridded climatie
data, land cover, and land use. Maps of GPP and NPP were del'ived from annually. integrated llormali~ed difference vegetation
index (NDVI) from the Moderate Resolution Infrared Speetroradiometer (M ODJS) and the Advanced V <lry B igh Resolution
Radiometer (A V HHH) satellite data. Climatic: data, particularly lIlean annual potential evapotranspiration [M APET ,Zomer 1"1
al . (2006)] ,was derived from the WORLDCLIM dataset developed by Bijmans et al . (200S) .
Secondly, a Dryland extent map was generated using gridded population, transportation data, e.g., data of roads and
railways. power consumption ,land use/land cover ami aridity index (AI) data in a subtractive overlay GIS procedlll'f' . A I is the
[nitI'd Nations Convention on Des"rtification's definition of Drylands where AI is the ratio of m"an annual precipitation (MAP)
to M APET that is ~;.:~ 0 .65 .
Thirdly, the N ASO data is used to generatl' livestoek numher and distrihul.inn maps which are intersee!ed with the drylaml
extent map to yield a livestock footprint map. Lastly, this footprint map is then converted to an annual requirement of forage
map anrl.ubtl'acted from a GPP map to yield a livestock appropriation of net primary productivity map (LANPP)
.References
Hijmans , R J .. S E . Cameron, J .L . Parra. P .C . Jones find A . Jarvis, (2005) . Very high resolution interpolated climate
surfaces for global land areas. International Journal of Climatology 25 : 1965-197g .
Washington-Allen, R J\. . ,R . D . Ramsey, N . E . West, and R . A . Efroymson . (200G) . A Remote sensing-based protocol
for assessing rangeland condition and trend. Rangeland Ecology and M. anagement 59 :19-29 .
Zomer • R . J ., Trabucco, A . , van Straaten .0 . ,and Bossio ,D . A . (200n) . Carbon, land and water: A global analysis of
the hydrologic dimf:'nsions of climate change mitigation through afforestation/reforestation and the Kyoto Protoeol Clean
Development Mechanism. Columbo, Sri Lanka: International Water Management Institute. pp 48 . IWM I Research Report
lOJ .
GrassJandS/Bangtiands Resources and Ecology
m m •••
Indicators for Sustainable Use and Conservation of GrasrJandS/Bangtiands Resources
Procedures for Estimation of the Livestock Ecological Footprint in US Drylands
Robert A. Washington-Allen 1 and John E. Mitchell
2
1. Department of Ecosystem Science and Management, Texas A&M University, 2138 TAMU, College Station, Texas 77843-2138, washingtonallen@tamu.edu
,
~qement
The recent Millennium Ecosystem Assessment of Orylands concluded that the amount of desertification ajti!f191
actions contribute to this process is unknown at national to global spatial scales.
Our research is relevant to this issue and conference because it seeks to develop procedures for determining the impact or ecological footprint of livestock grazing on the productive capacity of drylands at these spatial scales. The
study approach has been to develop a Geographic Information System (GIS) consisting of digital maps of gross and net primary productivity (GPP and NPP), Food and Agriculture Organization (FAO) or other sources of national
agricultural statistics on livestock number and distribution, transportation and power consumption networks, gridded climatic data, land cover, and land use. For example, for the United States, maps of GPP and NPP were derived
from annually integrated normalized differen-ce vegetation index (NOVI) from the Moderate Resolution Infrared Spectroradiometer (MODIS) and the Advanced Very High Resolution Radiometer (AVHRR) ntellite data. Climatic data,
particularly mean annual potential evapotranspiration IMAPET], was derived from the WORLDCLIM dataset. Secondly, a Oryland extent map was generated using gridded population, transportation data, •. g., data of roads and
railways, power consumption, land use/land cover and aridity index (AI) data in a subtractive overlay GIS procedure. AI is the United Nations Convention to Combat Desertification's definition of Orylands where AI is the ratio of
mean annual precipitation (MAP) to MAPET that Is:S: 0.65. Thirdly, the NASO data is used to generate livestock number and distribution maps which are intersected with the dryland extent map to yield a livestock footprint map.
Lastly, this footprint map is then converted to an annual requirement of forage map and subtracted from a GPP map to yield a livestock appropriation of net primary productivity map (LANPP).
Definition of Drylands
INTRODUCTION
Drylands make up 3,902,000 km· 2 of the United St.tes land surface and cover some 41% of the world's terrestrial surface whilst
providing ecosystem goods .nd services to some 38 % of the human population (M~lenium Ecosystem Asa.essment (MEA) 2005).
The ecological condition and trend of these lands is unknown.t regional, nationa,l.nd global scales (The H. John Heinz Center
2002, West 2003, Millenium Ecosystem Assessment (MEA) 2005, Washington.AJlen et 8/. 2006). Large geographic areas of dryland
are unmonitored at regional and national scales. Major problems are that dryland condition is measured in different ways by different
responsible agencies at local Icales and thus subect to different interpretations. Further water·limited systems are driven to some
extent by climatic events such as the EI NU\o-Southern Oscillation (ENSO) and La Nifta events that have retum intervals of between
5 to 10 years. Experimental design calls for replication of treatments, in this case both grazing and climate change variables.
Consequently, monitoring and assessment requires time Ierles of 15 to 2()..years of data or longer to detect trends. Additionally
local·scale data of an appropriate time series cannot be aggregated to satisfy national assessments. Nonettleless, public and
governmental responses from previous assessments have shown that interest in dryland condition is paramount. However, until
federal and national agencies .nd NGOs .dopt consistent indicators of ecological conditions that are broadly monitored, a different
approach is needed. We consider two indicators of human impacts on the face of the earth: the ecological footprint and
appropriations. The ecological footprint measures how much land and/or water area 8 particular population requires to produce the
:m;.~fIV~~umes and to absorb Its wastes. Appropriation considers the rate or the intensity of use of ttle world's photosynthetic
The extent of
~vestock
impact in drylands is the goal of this study. However, there are over 100 definitions of drylands that can be
used to set the extent. Fortunately, the United Nations Convention to Combat Desertification (UNCCD) had defined drylands globally
as the ratio of mean annual precipitation (MAP) to mean annual evapotranspiration (MAPEl) called the Aridity (or Humidity) Index
(AI) that i5.$0.65 (Figure 3). We will use the AI index developed by (Zomer fit aI., 2006) that was derived from the WORLOCLIM
dataset developed by Hijmans
~~b.:!':'}:. .~":;~t.~'<'~:f~~::..;~.~~.;'~~~~=. .":,",,,"''':''~~-
Thus the objectives of this research are:
1.To further develop and apply a protocol for remote.sensing based monitoring of U.S. drylands (Washington·A1len et 8/. 2006).
2. To assess trends in productive capacity of U.S. drylands using aspects of both an ecological footprint (Sanderson at al. 2002)
and an appropriations approach (Imhoff et aI., 2004).
METHOD
Our approach is to use a Geographic Information System (GIS) based approach to look at the spatial distribution of the intensity of
livestock grazing at ttle natlonal areal extent of U.S. drlyands In relation to dryland productivity. Specifically, this study will focus on
the year 2002, because this was when national statistics 01'10 both the number and spatial distribution of livestock across the United
States became publicly available.
Figure 3, Ttle global extent of Dryl.nds . . defined by the .rldity Inex (AI) which covers 41% of ttle terrestrial surface.
Database Development
Table 1 .how. the GIS data ba. . and m.p layers that will be developed to further renne the extent of dryl.nd. used by
r.ngeland livestockto produce. conslatent mettlodology to ...... trends In productive c.paclty of U.S. r.ngelancls.
"' __ cwr, .......
.~~~
"
[:2:~:
Livestock Distribution
Ttle National Agricultural Statistics Service (NASS, http://www.nass.usda.govf) conducts twiee.-yearly livestock surveys for the USA
at the county..JeveI from 1869 to the present. The US Census Bureau also collects agricultural data, including livestock data, that is
used by NASS to spatially represent US livestock distribution (Figure 1A and B). These number can be converted into forage
required or YDd by livestock.
A
B
""Cowa·lnnnlory: 1teZ
Sh.'pnL. . . ·k,..enl:CN)': 1"2
:r~:
....-
~. . . . . I:..:...
RESULTS
r---------------------~
Av.ll.blllty
Figure 1, Dot m.ps of the .pd.1 distribution of beef cow. (A) .nd .heep and lambs (B) across the US. Each dot
repreeents a ttl ....hold value of~5000 « 5000·0) bHf
or .heep .nd lamb•• The number of dots within a
p.rticul.r I.nd cover w . . b.nd on we~htlng factor. that were likelihood estimators of each type of livestock
occurring in • partlcul.r I.nd cover. The determined number of dots were then r.ndomly dl.tributed within the
.pecinc I.nd cover polygon (http://wwwJ.l.... u.d ••gov/re.e.rch/atl ••02latl ••-Info.html).
cow.
n
u..
Refined Drylands
Prod uctivity
B
A
Net Primary Productivity (NPP) marks the first visible step of carbon accumulation; it quantifieS the conversion of
atmospheriC CO2 into plant biomass (Running et at 2004). Consequently it is an indicator of productive capacity. NPP can
be derived using satellite remote sensing data, The Advilnced Very High Resolution Radiometer (AVHRR) is a remote
senor on board a satellite platform that has been globally collecting electromagnetic spectrum (EM) data since the mid·70's.
It coUect& data twice daily in. number of discrete sections or bands of the electromagnetic spectrum (EM). The infra..,.ed
(IR) .nd red (R ) bands of the EM are related in what is caned the normalized difference vegetation index (NOVI) which is
cak:ulated as:
NDVI-IR-R/IR+R
The Moderate Resolution Imaging Spectroradiometer (MODIS) is a remote sensor aboard both the AQUA and TERRA
satellite platforms which has been collecting global data since 1998 including NDVI and NPP. The NOVI has been
correlated to various indMdual plant (or leaf) .nd vegetation parameters including leaf area index (LAI), biomass, and
canopy cover (Sellers, 1985). Seasonally or annually integrated NOVI orrNDVI over the year has been found to be
correlated with NPP (Goward et al. 1985). Figure 2 shows an example of NPP for 2002 derived from the MODIS sensor.
NPP is considered
.
IlJI'Crc:t ..~)
OO'~'MI_
~
...
00 .. ·•
I:.:
Figure 2. Net Prim.ry Productivity fNPP) for 2000 derived from the Moderate Resolution Imaging Spectroraciiometer
(MODIS) ...,sor,
Figure 4, The Definition of dryl.nd. I. further refined using I.nd cover .nd urban distribution data fA). Thl. refined definition
I. then u.ed to bound the extent of dryl.nd. in term. of NPP (forage .v.n.ble) .nd consumption by livestock (forage use )
(B). Ttle result of the Intersection and union of the.e .re•• Is .n ntimate of the 1m of production impacted by livestock or
Livestock Appropriation of NPP (LANPP).
CONCLUSIONS
This initial study will look at data in 2002. However, future studies will both and predictively consider the time series of NPP and
livestock numbers and distribution data in order to retrospectively understand the consequences of impacts and to be able to
forecast their future impacts.
LITERATURE CITED
1. Goward, S. N., C. J. Tucker, and D. G. Dve. (1985), North American vegetation pattltml observed with the NOAA·7 advanced very high
resolullon r.dlometer, Vegelallo 64: 3 - 14.
2.H. John Heinz III Center for Science, Economics, .nd the Em/ironment (H.lnz Center). 2003. The st.te ofth. Nallon'. eco.ystem.:
Me.suring the land., wal.rs, and living resource. ofth. United St.t ••. The H. John Heinz III Cent.r for Science, Economics .nd the
Environment. New York, NY: C.mbrldge University Pre.s. 288 p.
Hijmans, R.J., S.E. Cameron, J.L. Parra, P.G. Jone••nd A. JaNis. 2005. V.ry high re.olutlon Interpol.ted climate swf.ce. for global land
are.l. Int. J. Climatology 25:1965·1978.
3. Imhoff, M.L., L. Bounoua, T. Rlcketta. C. Loucks, R. Harriss .nd W. T. Lawrence. 2004. Global patterns in human consumption of net
primary production. Nature 429: 870-873.
4. Millennium Ecosystem Anessmen! (MEA). (2005), Oryland Systems. pp. 623-662 In: MIII.nnium Ecosy.t.m ASHlSment, Ecosystems
Ind Human Well·belng: Cu".nt Stlte and r,.nds, Washington, D. C., I.land Pr.se.
5. Running, S. W., R. R. Nemenl, F. A. Helnsch, M. Zhao, M. Reeves, andH. H.llhlmolo. (2004), A contlnuou•••tellite-deliYed me.lllA of
global terreslrlal primary production, BioScience, 54:547 - 560.
6. Sanderaon, E. w., M. Jalleh, M. A. Levy, K. H. Redford, A. V. W.nnebo, .nd G. Woolmer. 2002. Th,bymanfoolprlnt Indth'I'st gfth.
wild. Blo&clenc. 52:891 .904
7. Sell .... , P. J. (1985), C.nopy rafltct.nc., photosynthesis .nd transpiration. Int,m.tlon.1 Joumlll of R.mot, S.nalng 6:1335-1372.
Washingion-AAen, R.A., R. D. Ramsev, N. E. West, and R. A. Efroymson. 2008. A remota ..""ng·bued protocol for ...Hllng rtngel.nd
condition .nd trend. Rangel.nd EcoL Manage. 59:19-29.
8. W.st, N. E. 2003. Theoretical underpinnings of r.ngel.nd monitoring. Arid L.nd R.... rch .nd Management 17:333-346.
9. Zomer, R. J., Trabucco, A., Vln Stra.t.n, 0., .nd Bo.sio, D. A. 2008. C.rbon, land .nd wat.r: A global .n./yll. oflhe hyd"ologlc
dimensions of climate ch.nge mitigation through affor.st.tlon/refor••tatlon .nd tne Kyoto Protocol CI•• n Oevelopmant Mechanllm.
Colombo, Sri L.nk.: Intern.tlon.1 W.ter M.n.gement lnatitute. pp.048. IWMI R....rch Report 101.