MANAGEMENT STRATEGIES
AND CLIMATE IMPACT CHANGE ON RANGELANDS'
JOSÉ DE SO1JZA NETO2, JAMES RICHARD CONNER5 , JERRY WAYNE STUTH4,WAYNE TRAVIS HAMILTOM and JAMES WILLIS RICHARDSON5
ÀBSTRACT - A hydrologic-based forage production simulation model (PHYGROW) and a popula-non mixture simulation model (POPMIX) were uscd respectively lo simulate forage producrion and carrying capacity of a subtropical shrubiand complcx of over 34 species grazed by various ratios of cattle and goats with a popuiation of indigenous animais (white-tailed deer) over a 20 year simulated weather wofile. The Farm Levei Income and Policy Simulation Model (FLIPSIM) were used to evaluate and quannf' the impacts of alternativa rnanagement strategies and clirnate change on grazingland ecosystems. The study was cariied out to analyze thc feasibility and profitability of a representative cante goal farm in lhe South Texas. On average, nel cash farm ineome under 50:50 cattle:goat ratio and climatic condínons falis by as much as 2% relativa to 70:30 for lhe farra studied. Real net worth for lhe farm declines as much as 9.4% and 16% over the study period under the highest to lowest cattle:goat ratio and, dry to normal climatic conditions, respectively. The modeled results produced useful information showing the socioeconomic consequences for a typical South Texas farm im-pacted for alternativa management strategies and climatic conditions.
Index terms: símulation, economic impacts, grazingland ecosystem
IMPACTO DE MUDANÇAS DE ESTRATÉGIAS DE MANEJO E DO CLIMA EM PASTAGEM NATIVA
RESUMO. Modelos dc simulação de forragem (PHYGROW) e de simulação de populações mistas de animais (POPMDC) foram usados respectivamentc para simular produção de forragem e estimar a capacidade de suporte numa complexa área de pastagem nativa com mais de 34 espécies de plantas arbóreo-arbustivas na qual pastcjavam bovinos e capnnos com uma população de animais nativos (veado) num período de 20 anos. Utilizou-se um programa de simulação em uma propriedade (FLIPSIM), para avaliar e quantificar os impactos de estratégias de manejo alternativas e de mudanças climáticas no ecossistema pastagem nativa. O principal objetivo do estudo foi ode unir os modelos já desenvolvidos. Com este procedimento integrado, o estudo procurou analisar a viabilidade e lucratividade de uma fazenda representativa da produção de bovinos e caprinos no Sul do Texas. Em média, a renda líquida da fazenda, com a relação bovino:caprino 50:50 e condições climáticas, cai cerca dc 2% em comparação com a relação 70:30 na fazenda estudada. O acervo de bens em termos reais referentes à fazenda declina cerca de 9,4% e 16% no período de estudo, quando passou da mais alta para a mais baixa relação bovino:caprino, bem como quando passou da condição climática seca para a normal. Os resultados do modelo produziram informações úteis que mostram as conseqüéncias socloeconômicas numa propriedade típica do Sul do Texas, afetada por estratégias de manejo alterna-tivas e por condições climáticas.
Termos para indexação: simulação, impactos econômicos, ecossistema pastagem nativa.
'Accepted for publication on December 5, 1997.
Extracted from lhe lirst wnter's thcsis prcsented lo the Texas' A&M Univcrsily, Texas, USA.
2Econ. Agrícola, Ph.D., Embrapa-Ccntm Nacional de Pesquisa
de Agroindiatria Tropical (CNPAT) Caixa Postal 3761
CEP 6051-110 Fortaleza, CE. E-mail: jsnctcnpatembrapa.br
3Econ. Agrícola, Ph.D., Ranching Systcma Group, Department ofRangcland Ecology and Managenieni, Texas A&M Univer-sity, Coilega Station, Texas 77843-2126 USA.
4Rec. Naturais, Ph.D., Ranching Systems Group.
3Rec. Naturais, M.Sc., Ranching Systems Group.
INTRODUCTION
The ordinary problems ofmanagmg rangelands chailenge lhe capacity of contemporary decision makers, institutions and scientists to integrate eco-logical, economic and social information.
Ecological modeling of climate change effects on rangelands fails to account for direct human ef-fects resulting from various factors such as popula. tion growth, economical activities, technology and
policy. On the otherhand, socioeconomic modeling tends to focus on the quantifiabte factors (eg., popu-lation market, demand, etc.) while negiecting the ecological and climatic change (Frederick, 1994).
The first step toward better integrated assessment ofthe effects of climate change on rangeiands should be linked with existing modeis. Thus, new modei-ing approaches are needed to better deal with lhe biophysical uncertainty and compiexity of the im-pacts oftechnoiogical and climate change on range-lands (Conner, 1994).
A substantial portion of iivestock production ia the world takes place ou grazinglands with and or semi-and climates (Holecheck et ai., 1989). Conse-quently, these areas are sensitive to changes ia the environmental conditions with which they are asso- • ciated, and also to climatie variations and changes in land use.
There are severai reasons for suspecting that rangelands would be very sensitive to the interac-tive effects of climate change. First, the quantity and seasonal distribution ofprecipitation is very impor. tant in determiningthe managementproductivity and distribution ofrangelands. Second, reduction ia pre-cipitation associated with increase in temperature may cause acceierated human-caused degradation ofecosystems. This range deterioration, a reduction ofthe range forage production potential, is a conse-quence of compiex processes that negatively alter plant cover, composition, and/or soil characteristics ofrangelands. Third,the deterioration ofrangeiands have occurred in a relatively recent period of time that coincides with their increased utiiization for lhe production ofdomestic animais (Parton ei ai., 1994). Despite the potential importance ofsocioeconomic impacts of climate change on rangeiands, current assessments ofthese impacts ou range lands are lira-ited by the lack ofadequate integrated modeis.
Currently, agricultural models are used to set
mi-
- tial conditions for the economic modeis, or specific management practices are seI as a background for the agricuitural modeis. The application ofdynamic vegetation growth models has not been caris idered in these modeis (Souza Neto, 1996). Thus,it
can be hypothesized that advances in integrating ecotogi-cal and socioeconomic modeis can be instrumental in developing the capability to assess the impact of cliniate change on rangelands.Pcsq. agropec. bras., Prasilia, v.33, n.9, p.1533-154l. sei. 1998
This study addresses this hypothesis and presents an integrated approach for tinking ecologicai and socioeconomic models in an innovative way that can help policy analysts and policy makers sort through the numerous impacts assessments, and theory ex-tend their information base. The idea was to inte-grate the use ofthe Farm Levei Income and Policy SimulationModel - FLIPSIM (Richardson & Nixon, 1986) with a Phytomass Growth Simulation Model - PHYGROW (Texas A&M University,
1995)
and an animal production and forage use simulator(Popu-iation Mixture Simulation Model - POPMLX ), a com-ponent decision tool in the Grazingland Appiications -GLA (Stuth et ai., 1990). This is the basis for this study.The central purpose ofan integrated assessment model isto organize compiex technicai information across disciplines. This study aimed at anaiysingthe feasibility and profitability of a representative farm facing natural events and influences under different climate and enterprise mix scenarios.
MATERIAL AND METHODS
Data were assembied for a typical South Texas farm comprised of 30% clay loam and 70% sandy loam range sites. Tbirty eight criucai piant species or functionai groups were idcntified in Lhe arca of study. Experts provided growth, preference and hydrological attributes of each specieslfunctional group. Vegetation survey of the La Copita Research Área - Texas was used to determine de Witt relative yield values for each specie/functional group for model piant communities ia Lhe region. Reiative yield value are the proportion (%) of fuli potential occupancy of a single species ar functional group on a piant commu-nity.
Sou
11ayer characteristics ofmodel sou 1 typical ofeach range site were assigned based on predominate acreage avaiiable in soil survey of the region. Ia order Lo generate a iong-term historical data seI lo devetop weather param-eters, Lhe USCLIMATE (1-lanson et ai., 1994) was geo-graphically adjusted for eievation, annual precipitation and latitude/longitude at the La Copita Research Arca. The soil laycr characteristics, surface features speies/functionaI group, plant communities, relative yieid (Tables 1 and 2), and weather generator parameters were cntered into PUYGROW lo generate forage production. The inventory of available forage (simuiated) using two different types of sou, clay loam and sandy ioam, was combincd to create data to be used ia the POPMIX. ThenMANAGEMENT STRATEGIES AND CLIMATE IMPACT 1535 j -i o, E 1., o - •0 j 1- E o o o E E E
a
2 2o 'e o, 2 f ri i o e OriNN'e'OO O 'oQ ri -r4,iri ri ri - - ri --- . 1' •, . . . •0• '1 11"1'1 o o c_c - ' o,III 1 1 1. 1 1." 1 111 liii Eoob.
1 III 1 1 1 1 1l 1 111 11111 'o o E 'o o 'o ri e o e e ri e O o e e e o E 0 - .a'ori eo----d - o 'o ° 'ir-.o. e'er-ecie — 0000 rio,.nri—OIO'o o eeee e ri — 'e'e'e - _uc E E
IH
-
H
• • _9-; ..i, o ._ °--e
• E -= •'o= 'o E • euis is 1 P o'o O ri ri E o E O'COO so E 44 ri — —————— — — — — — — — — — — — — --- íz E 000 o o o o o o o o 000000000 rirjriri.s —
hI
sW « '5 r5 rs ri ri ' . ri ri ri .2 .2 ri 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1111 111.1 1 lI 1 111111111 1 E .550.55.55——O r r 00000 000000000 o E ro E ri 2 2 °o .50 E E 2 ri °o 2 ri 2 O O OLI esoo: .1iri— ——O.005.j ...i
-- -- - - - --- - -- - - - -00— 00 . O o E _E rioov—-risoriri s.s ri ri ri ri ri ri ri ri ri ri ri ri ri ri ,1 ri ri — riOri .50'5 'eifç-1r1 o o o o o rj — _oo E •'° B . E
a
.'MANAGEMENT STRATEGIES AND CLIMATE IMPACT 1537
a transect production was estimated, and lhe forage (dry weight) for each specie was auached to POPMIX. The results of stocking rates provided by the POPMIX modei were used to generate basic input scenarios to be used in the FLIPSIM. Two cattle:goat ratios, 70:30 and 50:50 were selected to represent alternative management strategies aval lable to a farm operator.
Twenty years of simulated forage production and associatcd stocking raLes from POPMIX (Table 3) were divided into two ten-year periods; one representing nor-mal conditions scenario with 30% drought years and lhe other representing dryer ciimatic scenario with about 50%
drought years. Normal ycars are considered to be those that have Iess than 20% deviation of lhe long term average stocking rate, output from POPMIX, of a 20-year simulation. The stocking rates for the four ten-year period scenarios were used with aset of decision rules prescribcd by range animal scientists and ranch managers familiar with cante and meat goat enterprises in the South Texas arca to estimate animal production leveis and deviations from normal (average) annual operating cost for the two enterprises. The input ofPHYGROW and POPMIX will create lhe basic scenarios for lhe representative farm, for which economic performance wili be simulated by FLIPS1M. The annual animal production leveis and operating cost deviations were uscd as input data in lhe FLIPSIM. This whole-farm model is needed because it aflows the incorporation ofrisk and uncertainty, financial and accounting components, and institutional farm poli-eles into the calculations of economic performance ofthe farm.
The modeis work independently, and are controlied by lhe analist. The interrelations between lhe modais and the steps lo perform this study are included in Table 4.
Descriptive data for lhe representativa farm used in this study are summarized in Table 5. Information to describe lhe representative farm for the FLIPSIM model was obtained from budgets deveioped by extension farm ranch management speciaiist in the region of inlerest.
In this study, the representative farm represents average management using sound, accepled production and management practices. The South Texas cattic and goat farm has 493 goats and 466 cows (Tabie 5). The farm grazes 630 animal units on 5,000 ha ofrangcland owned. The combined stocking rale is 7.9 ha per AU. The initial value for lhe livestock on the farm was estimatcd at 5213,240.00. More complete detalis about lhe representative farm used in this study can be found in Souza Neto (1996).
RESULTS AND DISCUSSION
Two cattfe:goat ratios were selected based on trends in lhe region for expanding meat goat popu-lalions; 70:30 and 50:50. Although daily informa-tion of forage grbwth, species selecinforma-tion, and soll moisture were availabie, only annual production and consumption data were used in this analysis.
The 70:3 O and 5 0:50 ratios yielded stocking rates of 1.48 and 1.34 AUM/ha, respectively. The stan-dard deviation of stocking rate was 0.72 and
0.65 AIJM/ha, respectively for cattle. Goat stock-ing ratas and standard deviations for lhe 70:30 and 50:50 ratios were 0,119 (SD=0.06) and 0.15 (SD0.07), respectiveiy.
After running lhe 20-year analysis of the 70%
and 30% goat demand ratio, there were 9 "normal" stocking years for goats and 7 for caule (Fig. 1). Normal years are those with less than 20% devia-tion of the long-term average stocking rale of a 20-year simulation, Betow normal stocking years
(-20 to -5 0%) for cattie and goats comprised 25% of lhe years. Cattle would experience a higher percent of extremely tow stocking years than goats (15% vs. 10%). Ten percent of lhe years experi-enced stocking rates greater than 50% above nor-mal stocking leveis. Analysis ofSO:50 ratio resulted in similar trends.
A summary of selected indicators of the eco-nomic and financial condition of lhe representative farm that would be expected after ten years under each of the four prescribed scenario can be found in Table 6. Under normal weather scenario, the 70:30
catt!e:goat ratio produces less decline in real net worth, higher average annual cash receipts and net income, but lower average annual returns lo assets and equity and higher net income risk index than does the 50:50 ratio.
Under the dryer weather scenario, declines in real net worth are greater, cash receipts, net incomes and returns to assets and equity are lower regardless ofcattle:goat ratio than under normal weather sce-nario. As in the normal weather scenario, however, the 70:30 cattle goat ratio produces higher receipts and net income risk index than the 50:50 ratio.
— •0 ' - W 0% %O r 00— r4 %0 r. r 00 ' 00 e %0f. —v• I- -t 00%0 00 Q 0000 O O O O 0 O O O O O O O O O O O O O O O O O • % t' 00 4 0% - 0% 00 O e Q O V ri
E
2 ri ri e e 0% ri O O 00 0% ri 00 00 - ri ' 00 00 ri t. FS ri O W 00 O Sf1 — ri 0 e - ri 00 m e f r- 00 O q. — ri •- s e Sf1 Sf100 ooe 51) Q% ri 00 ' Sf1oO'O c - — 0Ci —0 - O O —O r1 0 — O ri 'o oo—— — 0% - 0 000 r- e — kn —O' 'O - VI o e i r- O 'O 'O ri 00 ri ri Q% ri O ID r e O e VI - t- •rs. e 0% e e 00 — 00 - 0000000000000000 O O O O 1 00 — CD ri ri e 0000 -.00 00 00000 VI — - riO 'oe p o ———————— o — — — 2 — O m o — o a — or sfl ri ri — 00 'O 'o 00 ri o O erieVI'O ooO'o ri VI%Ol-00O%O
Pesq. agfopec. bras. BraslHa, v33, n9. p.l 533-1541 • seL 1998
MANAGEMENT STRATEGIES AND CLIMATE IMPACT 1539 TABLE 4. Steps ia analysis oísocioeconomic impacta of change on rangelands using curreal modela'.
Step number Discrimination
1 Select representative farms for modeling.
Fai-ins should be representative ofresources, size and tenure ofproduction units and production practices in a specified portion of the study arca.
2 Run PHYGROW for each representative farm for each year in the 20-year period.
lnput data Output
a. dlimate a. forage production by type (grass, forb,
b. soils browse) by season, by year
c. hyclrology d. vegetation
3 Run POPMIX and AS? modeis for each representative farm for each year in lhe 20-year period.
Input data Output
a. initial animal inventories and characteistics a. animal
by kind and class and class (including b. forage production yicld and inventory vegetation preference rating) by kind and c!ass by year.
b. husbandry practices c. production goals
d. foragelfeed availability by type, by season, by year
4 Run FLIPSIM model for cach representative farm for each year in 20-year period under different scenarios.
Input data Dutput
a. beginning financial characteristies offarm a. end ofplanning period financial b. annuat institutional and macroeconomic characteristics offarm
parameters b. annual cash flow
c. annual enterprise production requirements c. probability offarm survival andlor (factors) and product and factor prices andlor growth over planning period. probability distributions
5 Repeat steps 2-4 and compare results under diffcrent scenarios to determine lhe impacts of change (climate, technological or institutional) on future socioeconomic characteristics of the representative farm and make inferences about region of interest.
1 Sourcc Adaptcd from Cooner (1994)
the South Texas'.
Characteristics South Texas cattic
goat farm
Total animal units (AUs) 630
Goats Nannies (no) 493 Replacement (no) 100 Bucks (no) 30 Cattle Cows (no) 466 Rcplacement (no) 72 BulIs (no) 12 Assets Land ($1.000) 5,559,750 Machinnery ($1.000) 78,336 Livestock ($1.000) 213,240 40 1' Total ($1.000) 6,117,942 Efficicncymeasure 1 3 5 7 O 11 13 15 17 19
Calving (%) 93 yui' lo plannjng horIon
Kid crop (%) 150 FIG. 1. Simulated annual deviation from 20-year av -
Steer weaning wcighc (kg) 250 erage stocking rate of a mixed livestock popu-
Goat weaning weight (kg) 16 lation (70:30; 50:50) cattle: goat ratio with a
resident herd population ofwhite-tailed deer
Source: Adapted from Souza Neto (1996). (4.86 ha/head).
TABLE 6. Implications for alternative stocking rate for a South Texas cattle and goata farm under two climate sceno rios.
Indicators Weather
Dry Normal
70:30 . 50:50 70:30 50:50
Change in Real Net Worth 1996-2005 (%) -2.65 -2.9 -2,12 -2.46
Cash Receipts 1996-2005 ($1.000) 220.57 211.54 225.97 218.37
Expenses to Reccipts 1996-2005 (%) 50.9 49.6 47.1 46.19
Net Cash Farm income 1996-2005 ($1.000) 110.72 108.6 122.19 119.75
Risk Index for Annual Net Cash Income (%) 14.95 9.96 2.00 O
Average Return to Asscts 1996-2005 (%) 0.48 0.57 0.50 . 0.59
Average Return to Equity 1996-2005 (%) 0.40 0.53 0.43 0.55
Source: Souaa Neto (1996).
Pesq. agropec. bras., Brasilia, v.33, n.9, p.1533-1 541, set. 1998
-70% c&ttls 120 80 40 30 1111141144411 1 II 11 -1 1 3 5 7 O 11 13 15 17 11 Y..r In planning horizon
MANAGEMENT STRATEGIES AND CLIMATE IMPACT 1541
CONCLTJSIONS Washington, DC: USDA. Agricultural Researcb
Service, 1994. IlIp. I. Higher ratio of goat to caUle stabilize annual
variation iii stocking as ecological conditions shift from a grassland domain loa shrubland domam and chmatic conditions become drier.
2. Goats whilc generaily less profitable than caule, rcqwre less annual operating cost, less capital iii-vestrnent and exhibit less variation iii annual receipts and net income.
3. Tlie lower variation iii net incomes from goats compared lo caUle is due both to lhe goats forage availability bemg less impacted by variations m an-nual precipitation and less year-to-year change in prices received for meat goats compared to weaned calves sold from the cow-calfcnterprise.
REFERENCES
CONNER, J.R. Assessing the socio-economic impacts of climate change on grazinglands. In: FREDERICK, K.D.; ROSEMBERG, N.J. (Eds.). Assesslng the Impacts of dlimate change On natural resource system. Boston: Kluwer, 1994. p.143-157. FREDERICK,, K.D. Integrated assessment ofthe impacts
of climate change on natural resources. In: FREDERIC, K.D.; ROSEMBERO, N.J. (Eds.). As-sessing the impacta of dimate change ou natural
resource system. Boston: Kluwcr, I994.p. 65-90.
HANSON, C.L.; CUMMING, K.A.; RICHARDSON, C.W. Mlcrocomputer program for daily weather
simulation In the contlnuous United States.
HOLECHECK, .JL.; PIPER, LD.; HERBEL, C.H.
Range management principies and practleu,
Englewood Cliffs, Ni: Prentice-Hall Inc., 1989. 50Ip.
PARTON, W.; SCHIMEL, D.; OJIMA, D. Modeling environmental changes in grasslands. In: FREDERJCK K.D.; ROSEMBERG, N.J. (Eds.).
Ássesslng lhe Impacta ofelimate change on nata-ral reource system. Boston: Kluwcr, 1994.
p.lIl-141.
RICHARDSON, J,W,; NIXON, C.J. Descriptlon of
FLIPSIM - V: A general f'irm levei policy
simula-tion model. College Stasimula-tion: Texas A&M Univer-sity, 1986. I35p. (Texas Agricultural Experiment Station. Bulletin, 1528).
SOUZA NETO, J. de. Integratlng ecological and eco-nomie modeis for assesslng institutional, environ. mental and technological change on rangeland.
Coliege Station: Texas A&M University, 1996. I84p. Ph.D. Dissertation,
STUTH, i.W.; CONNER, J.R.; HAMILTON, W.T.; RIEGEL, D.A.; LYONS, B.; MYRICK, B.; COUCH, M.A. Resource planning model for inte-grated resource management. Journal of Blogeography, v. 17, p.531 -540, 1990.
TEXAS A&M UNI VERSITY, Department ofRangeland Ecology and Management. Ranching Systems Group. PHYGROW: phytomass gro1b siinulator. Version 2, model documentation. College Station, 1995. p.92. (Minieo, 95.1),
