13652494, 2019, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/gfs.12445 by University Of Waterloo Dana Porter Library, Wiley Online Library on [02/12/2022].

See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
| |
Received: 18 March 2019    Revised: 15 July 2019    Accepted: 29 July 2019

DOI: 10.1111/gfs.12445

ORIGINAL ARTICLE

Grazing management for more resilient mixed livestock farming

systems on native grasslands of southern South America

Pablo Modernel1,2,3  | Valentin Picasso2,4  | Martin Do Carmo3  | Walter A. H. 

Rossing  | Marc Corbeels  | Pablo Soca  | Santiago Dogliotti  | Pablo Tittonell6
4 5 6 3

1
Farming Systems Ecology, Wageningen
University, Wageningen, The Netherlands Abstract
2
Agronomy Dept, University of Wisconsin- Droughts in southern South America affect grazing systems in many ways. They re-
Madison, United States
duce biomass productivity; decrease livestock feed intake, weight and reproductive
3
Facultad de Agronomía, Universidad de la
República, Uruguay
performance; increase farmers’ costs; and reduce farm income. It was hypothesized
4
Farming Systems Ecology, Wageningen that simple grazing management variables affect the resilience of grazing systems to
University & Research, the Netherlands droughts at the paddock and farm scales. The effects of grazing management on herb-
5
Agroécologie et Intensification
age and animal production were assessed at paddock level, and how technological
Durable (AïDA), Centre de coopération
Internationale en Recherche Agronomique and structural variables relate to the production and economic performances at farm
pour le Développement (CIRAD), Université
level. Results of a grazing experiment controlling herbage allowance at paddock level
de Montpellier, France
6
Agroecology, Environment and Systems showed that resistance of herbage accumulation and animal live weight to drought
Group, Instituto de Investigaciones was significantly higher for paddocks with higher pre‐drought herbage allowance than
Forestales y Agropecuarias de Bariloche
for those managed to low herbage allowance treatments. A strong positive linear re-
Correspondence lationship was found between pre‐drought herbage height and resistance of herbage
Pablo Modernel, Farming Systems Ecology,
Wageningen University, P.O. Box 430, accumulation rate (p < .01). In a longitudinal study of nine farms in Uruguay, resistance
Wageningen 6700 AK, The Netherlands. of cow pregnancy rate to drought was positively correlated with cow pregnancy rate
Email: pablomodernel@gmail.com
(r = .72, p = .02) and farm net income (r = .78, p = .02), and negatively correlated with
Funding information sheep‐to‐cattle ratio (r  =  −.80, p  =  .01). These correlations suggest that farms with
EACEA (Education, Audiovisual and Culture
Executive Agency) of the ; European higher incomes and low proportions of sheep in the herd withstand drought better (in
Commission; CSIC—UDELAR terms of pregnancy rate). Four common regional production strategies were identified
that react differently when farmers face drought, and these results can aid farmers in
those regions to design more resilient mixed livestock farming systems and can inform
[Correction added on 08 October after first online
publication: Complete list of affiliations is reflected policymakers about effective strategies for mitigating drought impacts in the region.
in this version]

KEYWORDS
drought, grazing management, livestock farming systems, native grasslands, resilience, Rio de
la Plata grasslands

1 |  I NTRO D U C TI O N precipitation is increasing, both at the global and local scales (IPCC,
2013). For instance, in southern South America, weather records
Climate change is a major challenge for food security (Douxchamps for the last 50  years reveal increased variability in rainfall patterns
et  al., 2016; Godfray et al., 2010). Historical climate records and (Barros, Clarke, & Dias, 2006; Marengo et al., 2012). Projecting this
future climate model projections show that variability in annual trend forward in time, more frequent water deficits are expected

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium,
provided the original work is properly cited.
© 2019 The Authors. Grass and Forage Science published by John Wiley & Sons Ltd.

636  |  
wileyonlinelibrary.com/journal/gfs Grass Forage Sci. 2019;74:636–649.
 13652494, 2019, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/gfs.12445 by University Of Waterloo Dana Porter Library, Wiley Online Library on [02/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
MODERNEL et al. |
      637

(Shiu, Liu, Fu, Dai, & Sun, 2012). As a result, it is increasingly import- al., 2003; Overbeck, Müller, Pillar, & Pfadenhauer, 2006; Soca et al.,
ant to investigate the resilience, resistance and recovery of agricul- 2008) and can be achieved through stocking rate management (Do
tural systems at multiple hierarchical scales (Grimm & Wissel, 1997; Carmo, Claramunt, Carriquiry, & Soca, 2016; Sollenberger, 2015).
López‐Ridaura, Keulen, Ittersum, & Leffelaar, 2005a; Picasso, Casler, Higher levels of aboveground biomass are related to deeper and
& Undersander, 2019). “Resilience” is a key concept in this investi- denser rooting systems, and these improve the resilience of grass-
gation, and this term refers to the ability of a system to withstand lands to droughts (Bartaburu, Duarte, Montes, Morales Grosskopf,
a short‐term crisis, perturbation or shock, like a drought (Grimm & & Pereira, 2009; Norton, Malinowski, & Volaire, 2016; Van Ruijven
Wissel, 1997; López‐Ridaura, Keulen, Ittersum, & Leffelaar, 2005b; & Berendse, 2010). As a result, on land with higher herbage allow-
Picasso et al., 2019); it can be studied by focussing upon two comple- ances, animal live weight is less affected during drought, and fewer
mentary attributes: “resistance”, which is the extent of change due farm inputs and fewer farm management economic expenses are
to a crisis, and “recovery”, which is the ability of a system to recover needed to withstand the drought (Cobon et al., 2009; FAO, 2013).
from a crisis (Oliver et al., 2015). However, there is a need to translate results generated from these
Grazing systems are highly dependent on rainfall, which makes experiments to management recommendations at the farm level
them vulnerable to extreme events like droughts. Droughts are (Briske et al., 2008; Teague, Provenza, Norton, & Steffens, 2009).
perceived by contemporary farmers as one of the most important Integrating management variables at paddock level (e.g., herbage al-
shocks stressing their production systems (FAO, 2013). They force lowance) with technological variables at the farm level (e.g., stocking
farmers to sell cattle at a low price, resulting in economic loss, or rate) and farm performance (e.g., beef productivity or farm income)
may cause the death of animals (Cruz et al., 2018), resulting in more is a challenge this paper aims to address.
serious economic loss. In response, it was proposed that a resilience‐ Here, the relationship between resilience to droughts and graz-
oriented approach may assist contemporary farmers to address the ing management practice at the paddock and farm level was studied.
threats of droughts through development of a knowledge system The effects of grazing management on herbage and animal produc-
that employs ecological models and science to improve on‐farm tion were assessed at paddock level, and how technological and
management (Bestelmeyer & Briske, 2012). structural variables relate to the production and economic perfor-
Drought effects cascade through grazing systems: they reduce mances at farm level. An investigation was made of the effect of
plant growth, biomass and primary productivity; decrease livestock grazing management on a series of resilience metrics at these two
feed intake, weight and reproductive performance; increase farm hierarchical levels where farmers make their management decisions.
production costs; reduce income for farmers; affect rural communi- Hypotheses were:
ties; and even affect nation‐scale economies (Ahmed, Azeze, Babiker,
& Tsegaye, 2002; FAO, 2013; Paolino, Methol, & Quintans, 2010). 1. at the paddock level, higher herbage allowance increases resil-
Appropriate grazing management offers an opportunity to buffer ience of grazing systems to drought; that is, grazing systems resist
the effect of droughts on biodiverse grasslands (Cobon et al., 2009; and/or recover faster after the drought in terms of grass and
Thurow & Taylor, 1999). The biodiverse grassland of interest in this animal productivity when managed at higher herbage allowance;
study lies in the Río de la Plata grasslands region of southern South 2. at the farm level, lower stocking rates increase the resilience of
America. This is a grassland biome with 400 years of livestock rearing livestock reproductive and productive parameters and economic
history that harbours more than 4,000 native species of C3 and C4 indicators to drought.
grasses (Soriano, 1992). While species composition is highly diverse
and heterogeneous, the most common grass genera are Poa, Bromus,
Stipa, Briza, Piptochaetium, Paspalum, Panicum, Bothriochloa, Digitaria
2 | M ATE R I A L S A N D M E TH O DS
and Setaria (Bilenca & Miñarro, 2004). In this region, cattle (mostly
Britannic Rare Breeds: Hereford and Angus) and sheep (mostly
2.1 | Study region and farming systems
Corriedale and Merino breeds) graze all year round. Overgrazing
(Carvalho & Batello, 2009; Gutierrez & Modernel, 2011; Maraschin, Uruguay is part of a large biome known as the Rio de la Plata grass‐
2001; Overbeck et al., 2007) and changing land use are the main lands, where native grasslands and cattle have coexisted for the
threats to this semi‐natural agroecosystem (Modernel et al., 2016). last 400 years. There is a high diversity of farming systems in the
Long‐term experiments in Brazil and Uruguay have shown that region (Modernel et al., 2018), but two main types of farms can
herbage allowance (kilograms of herbage dry matter available per ki- be distinguished: reproduction or “cow–calf” and meat produc-
logram of animal live weight) (Allen et al., 2011; Sollenberger, Moore, tion or “finishing” farms (Becoña, Astigarraga, & Picasso, 2014;
Allen, & Pedreira, 2005) is a key variable in grazing management that Modernel et al., 2016). This study focused on cow–calf systems,
can be regulated to significantly increase current productivity and which represent 53% of farming systems in the country (MGAP,
income (Claramunt, Fernández‐Foren, & Soca, 2017; Da Trindade et 2015). The main income of cow–calf farms comes from selling
al., 2012; Do Carmo, Sollenberger, Carriquiry, & Soca, 2018; Soares calves and culled cows. As a result, pregnancy rate is a key pro-
et al., 2003). Higher herbage allowance corresponds with a taller duction factor. Native grasslands typically provide 90%–100%
sward. This, typically, will have greater species diversity (Carvalho et of the diet of these animals. As complementary income, these
 13652494, 2019, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/gfs.12445 by University Of Waterloo Dana Porter Library, Wiley Online Library on [02/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
|
638       MODERNEL et al.

farms also raise sheep for wool or meat production, giving rise to van Keulen, et al., 2005b; Picasso et al., 2019; Tittonell, 2014; Van
mixed grazing systems where sheep and cattle compete for the Ruijven & Berendse, 2010). Here, effects of varying grazing man-
native grasslands as feed resource (Paparamborda, 2017; Ruggia agements on the productivity of grazing systems responding to a
et al., 2015). Finishing farms mainly fatten male calves that may perturbation (drought) were analysed. Two resilience metrics were
be fed on native grasslands, leys or grains (feedlots) (Modernel, selected for this purpose: resistance (Walker, Holling, Carpenter, &
Astigarraga, & Picasso, 2013). Precipitation in the Rio de la Plata Kinzig, 2004) and recovery (Holling, 1996; Urruty, Tailliez‐Lefebvre,
grasslands region is highly variable between and within years & Huyghe, 2016). Resistance was determined as the magnitude of
(Barros et al., 2006; Caffera, 2005) with recurrent spring–sum- the effect of a perturbation on the system (e.g., how much the pro-
mer droughts (Cruz, Baethgen, Picasso, & Terra, 2014). One of the ductivity decreased during the drought relative to a normal weather
most severe droughts in the last 30 years in Uruguay occurred in situation). This was calculated as the ratio between the state of the
the spring–summer of 2008–2009 (Cruz et al., 2014; Paolino et response variable during the drought and a previous normal season,
al., 2010) (Figure 1). for instance, herbage mass during the season of drought divided by
herbage mass during the same season of the previous year with nor-
mal weather (Figure 2a). Recovery was determined as the speed at
2.2 | Resilience metrics
which the system came back to a previous stable state after the drop
The concept and theoretical framework of “resilience” have evolved caused by the perturbation. It was calculated as the slope of the vari-
since Holling (1973) first defined it (Ives & Carpenter, 2007). Many ef- able against time after the perturbation, for instance, the increase in
forts have also been undertaken to operationalize resilience metrics in herbage mass between the season of drought and the season after
agricultural literature (van Apeldoorn, Sonneveld, & Kok, 2011; Cabel the drought divided by the period of time (Figure 2b). While resist-
& Oelofse, 2012; Darnhofer, Fairweather, & Moller, 2010; Groot, ance is unitless, recovery is expressed in the units of the variable per
Cortez‐Arriola, Rossing, Massiotti, & Tittonell, 2016; López‐Ridaura, unit of time (i.e., season or year).

F I G U R E 1   Thirty‐year average
seasonal rainfall (continuous line) and
2007–2010 seasonal rainfall (dotted
line) in Treinta y Tres (−33°15′S, 54°28′,
eastern Uruguay). Data from INUMET
(2018)

F I G U R E 2   Metrics used for the assessment of resilience. The behaviour of a response variable is depicted for two different systems 1
and 2 (grey and black lines). (a) Calculation of resistance (higher for system 1 than system 2); (b) Calculation of recovery (faster for system 1
than system 2). RVb: response variable before drought; RVd: response variable during drought; RVa: response variable after drought; tb: time
before drought; td: time during drought. ta: time after drought
 13652494, 2019, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/gfs.12445 by University Of Waterloo Dana Porter Library, Wiley Online Library on [02/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
MODERNEL et al. |
      639

measured at intervals ranging from 19 to 61 days between 2007 and

2.3 | Case studies
2010. At the beginning of each measurement period, herbage avail-
The resilience analysis was performed in the Rio de la Plata grasslands ability (kg DM ha−1) was measured in order to allocate the animal live
region in Uruguay for two case studies at two different levels: pad- weight required to fit the herbage allowance levels previously de-
dock and farm (Figure 3). fined. This means that the number of cows (and thus the stocking
rate) changed from one measurement period to the next. Detailed
experimental measurements and results were reported by Soca et al.
2.3.1 | Paddock‐level analysis
(2008) and Do Carmo et al. (2018).
An experiment to assess the effect of herbage allowance on na- Herbage mass and height were quantified using the comparative
tive grassland and animal performance was carried out at the “Prof. yield method (Haydock & Shaw, 1975). Ten reference quadrats were
Bernardo Rosengurtt” experimental station in Cerro Largo (Uruguay) used for calibration, and 100 randomly selected quadrats were rated
(32°35′S, 54°15′W, 160 m) between June 2007 and February 2010. at each measurement date on each paddock (Table 1). Herbage height
The experimental area comprised 92 ha, and the experimental design was measured in the quadrat, and standing biomass was cut to ground
consisted of randomized blocks, with two replicates in each block, and level to quantify herbage mass. Herbage accumulation rate was mea-
eight paddocks in total. Blocks were selected by soil type: clayey soils sured by means of exclusion cages with a surface area of 0.25 m2. As
(Argiudolls) and sandy soils (Hapluderts and Argiudolls). The manage- described by Do Carmo et al. (2018), in each paddock two areas of
ment variable at paddock level was herbage allowance. This metric is 0.25 m2 were selected with similar herbage mass. In one area, herbage
used in grazing experiments because it can be directly related to ani- height was measured, and herbage was cut to soil level and weighed
mal performance since it describes the quantity of herbage per unit of after drying at 60°C until constant weight. The second area was cov-
animal weight. Two treatments were applied: low (3.5 kg dry matter ered by a cage to exclude grazing until the end of the exclusion period
[DM] kg−1 live weight [LW]) and high (5 kg DM kg−1 LW) herbage al- (between 23 and 111 days). At the end of the exclusion period, herbage
lowance (hereafter referred as LHA and HHA respectively). The LHA biomass was harvested, dried and weighed in the same way as for the
treatment mimicked the typical management in most of the farms of first area. The difference in herbage mass between the sampling dates
the country (overgrazing), while the HHA treatment aimed to match was defined as herbage accumulation (’T Mannetje, 1978). Cows were
the herbage availability with animal nutritional demands. Herbage ac- weighed at the beginning of each exclusion period, and the weight was
−1
cumulation rate (kg DM ha day ), herbage mass (kg DM/ha), herb- averaged per animal per paddock. The animal live weight gain was cal-
age height (cm), stocking rate (kg LW ha−1 year−1), cow body condition culated as the difference of the animal weight of each animal between
score, cow weight (kg LW) and calf weight at weaning (kg LW) were consecutive periods, divided by the number of days of the period.
Resistance to drought was calculated for herbage accumula-
tion rate and animal live weight gain as the ratio between the av-
erage of spring–summer 2008–2009 (drought) and the average of
spring–summer 2007–2008 (before drought) (Figure 2). Recovery
from drought was calculated for each measured variable as the slope
of the regression line between the averages of the spring of the
drought year and the winter after the drought.

2.3.2 | Farm‐level analysis
Longitudinal data of nine livestock farms included productivity re-
cords from at least 3 years before and after the drought of the spring–
summer of 2008–2009 and were collected by the National Livestock
Extension Agency in Uruguay (Instituto Plan Agropecuario). These
farms were selected since they are broadly representative of the
typical farm types of the region, and have reported productive and
economic information for 6 years. The farms are located in two re-
gions, one in the north (Basalt) and the other one in the East (Eastern
Sierras) of Uruguay (Modernel et al., 2016) (Figure 3), and manage
mixed beef cattle and sheep, grazing on native grasslands (Table 1).
The two regions were selected because the land use in both regions
F I G U R E 3   Location of the case studies in Uruguay. The
was predominantly native grasslands (more than 80%) and because
resilience analysis at paddock level was done with data from the
they encompassed most cattle farmers in Uruguay (65%). The varia-
experimental station. The analysis at farm level with farm survey
data from the North (Basalt) and East (Eastern Sierras) regions. bles studied on these cow–calf operations were cow pregnancy rate,
(grey shadow) equivalent meat productivity and farm net income (gross income
 13652494, 2019, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/gfs.12445 by University Of Waterloo Dana Porter Library, Wiley Online Library on [02/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
|
640       MODERNEL et al.

TA B L E 1   Sampling methods, mean, coefficient of variation (CV), minimum (Min) and maximum (Max) values for the studied variables at
paddock level

Variable Samples per paddock Method Mean CV (%) Min Max

Herbage height (cm) 10 reference quadrats Comparative yield method 6.3 82 0.5 35.6
Herbage biomass (kg DM for calibration; 100 ran- (Haydock and Shaw, 1975) 3,017 79 160 16,444
ha−1) domly selected quadrats

Herbage accumulation rate 32 exclusion cages of ’T Mannetje (1978) 13.4 103 0 85

(kg DM ha day−1) 0.5 × 0.5 m
Animal weight (kg) 1–24 animals Individual weight measurement 438 14 307 602

minus direct costs). The equivalent meat productivity is a metric

3.1.2 | Drought resistance and recovery
used in Uruguay to compare different production systems with vari-
able proportions of sheep and cows in the herd. The metric trans- Resistance of herbage accumulation and animal live weight to drought
forms the energy of beef, wool and lamb into a single unit (beef kg was significantly higher for HHA than LHA (p < .01) (Table 4). Herbage
LW + lamb kg LW + 2.5 × kg wool) (Oficialedgui, 1985). Resistance accumulation rate during the drought was on average 66% and 46%
of pregnancy rate and productivity to drought was calculated as the of the rates in normal spring–summer seasons for HHA and LHA re-
ratio between the drought year (2009) and the average values of spectively. While this reduction is an immediate problem during the
the farm during the previous 1–3 years, depending on data availabil- drought period, it also has longer lasting effects on herbage avail-
ity of each farm. Recovery of pregnancy rate and productivity from ability, since spring–summer is the season when most of the herbage
drought was calculated as the slope of the regression between the accumulation occurs (Berretta, Risso, Montossi, & Pigurina, 2000;
year of the drought (2009) and the next year (2010). Risso, Ayala, Bermúdez, & Berretta, 2005; Soriano, 1992). Animal live
weight during the drought was 98% and 94% of the weights in normal
spring–summer seasons for HHA and LHA respectively.
2.4 | Statistical analysis
Recovery of herbage accumulation rate was on average 3.37  kg
Analysis of variance (ANOVA) was used to evaluate differences in DM/ha day−1 season−1, representing an accumulation of 306  kg DM
resistance and recovery between herbage allowance treatments at in one season, not different between herbage allowance treatments
paddock level. Correlation matrices were calculated for all variables (p < .05, Table 4). Recovery of annual live weight was on average 19.4 kg
(at paddock and farm levels, respectively) to explore relationships
between management variables and output variables. All analysis
was performed using the “stats” R package (R Core Team, 2015).

3 |  R E S U LT S A N D D I S CU S S I O N
3.1 | Paddock level
3.1.1 | Herbage accumulation and animal weight
The herbage accumulation rate was the lowest in winter and highest
in spring and summer respectively (Figure 4a). The effect of drought
is more evident from the lower values in the spring–summer of 2008–
2009 than in the prior and posterior spring–summer seasons. Animal
live weight changed in a similar way to the herbage accumulation rate
(Figure 4b). Animals lost weight during winter and gained weight in
spring–summer before and after the drought. During the spring–sum-
mer of 2008–2009, animals lost weight due to the drought‐induced
reduction of the herbage accumulation rate, the herbage mass and,
consequently, the herbage feed intake of the animals.
High (HHA) and low herbage allowance (LHA) treatments gener-
ated different responses to drought. Animal weight was higher for HHA F I G U R E 4   Herbage accumulation rate (a) and animal live
weight (b) for high (HHA) and low (LHA) herbage allowance
than LHA (p < .05) in all the measurement periods after spring 2008 and
treatments. Points indicate average values per season. The black
over the entire experimental period (Table 2). Similarly, herbage height
line corresponds to high herbage allowance; the grey line indicates
and accumulation rate were significantly higher for HHA than LHA over low herbage allowance. The grey shadow area indicates the drought
the 3‐year experimental period (Table 2) (Do Carmo et al., 2018). period. Adapted from Do Carmo et al. (2018)
 13652494, 2019, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/gfs.12445 by University Of Waterloo Dana Porter Library, Wiley Online Library on [02/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
MODERNEL et al. |
      641

TA B L E 2   Mean, coefficient of variation (CV), minimum (Min) and in paddocks managed with LHA, herbage accumulation rate during
maximum (Max) for selected variables of the studied farms drought ranged between 35% and 55% of average productivity be-

Variable Mean CV (%) Min Max fore the drought. Pre‐drought herbage height is therefore a good
predictor of grassland resistance to drought (Figure 5), and LHA pad-
Soil productivity indexa 103 64 58 270
docks clearly suffered greater reductions than HHA paddocks.
Grazing area (ha) 621 55 157 1100
Vogel, Scherer‐Lorenzen, and Weigelt (2012) found similar results
Area improved pastures (%) 10 77 0 23
in an experiment designed to measure the effect of management on
Cattle stocking rate (LU/ha) 0.53 29 0.31 0.77 resistance to drought. They artificially simulated a prolonged sum-
Sheep stocking rate (LU/ha) 0.26 78 0.02 0.57 mer drought in plots with different: numbers of grass species (1–60)
Total stocking rate (LU/ha) 0.81 21 0.62 1.17 and management intensity (four levels of mowing frequency and
Sheep‐to‐cattle ratiob 2.9 88 0.1 7.6 fertilizer doses). Their main conclusion was that moderate grazing
Cow pregnancy rate (%)c 74 11 64 84 intensity with higher herbage height improves the resistance of the
Meat productivity (kg/ha) d
87 21 65 121 grassland when faced with a drought.
Net income farm (US$/ha) e
19 88 −65 98 Our field observations confirm that weather forecasts and early
f warning systems provide farmers with relevant information needed to
Input‐to‐output ratio 0.79 73 0.29 3.4
face an upcoming drought and to make mediating decisions (Ahmed et
Abbreviations: LU, livestock unit (a cow of 380 kg).
a al., 2002; Bestelmeyer & Briske, 2012; Cruz et al., 2018), and our ex-
‘CONEAT’ index, used in Uruguay as a proxy for soil quality (MGAP,
2018). The average index for the country is 100. periment suggests that the pre‐drought status of the herbage mass/
b
In livestock units. height should be taken into account. According to our results, an in-
c
(Weaned calves/served cows) × 100. crease of one cm in herbage height reduces the impact of drought on
d
In equivalent meat productivity (beef kg LW + lamb kg LW + 2.5 × kg
herbage accumulation rate by 20% (Figure 5). Managing grazing sys-
wool).
e
Gross income – direct costs. tems to HHA seems to give farmers and farms the capacity to better
f
Ratio between costs (input) and income (product). maintain herbage productivity and to better withstand the drought.
Differences in resistance between HHA and LHA were found
−1 for animal live weight (Table 3). These indicate that animals were
LW season , representing gains of 0.21 kg LW/day after the drought,
not different between herbage allowance treatments (Table 4). This able to withstand the drought and lose less weight under HHA. This
suggests that regardless of the grazing management methods, graz- could be a result of either the higher pre‐drought herbage mass (rep-
ing systems in native grasslands can recover quickly when rainfall resented by height) of the HHA treatment or the higher accumula-
resumes after drought. While this experimental result suggests the tion rate observed during the drought in the HHA. Both allow for
good resilience of Uruguay's grasslands, we caution that this condition greater herbage intake in HHA during the drought. This observation
might change under conditions of long‐term overgrazing (under LHA). confirms other studies, whose overall results show that higher HHA
Conditions that might affect resilience include the appearance of areas increases animal productivity as a whole (Claramunt et al., 2017; Do
of bare soil, induced soil erosion (Roesch Wurdig et al., 2009) and in- Carmo et al., 2018).
vasion of non‐productive non‐native species (such as Eragrostis plana
Nees) (Bresciano, Rodrigues, Lezama, & Altesor, 2014; Focht & Borges
3.2 | Farm level
de Medeiros, 2012). These were not, however, observed during our
experiment.
3.2.1 | Meat productivity, cow pregnancy rate and
A strong positive linear relationship was found between pre‐
farm income
drought herbage height and resistance of herbage accumulation rate
(p < .01, Figure 5). Resistance, or the ratio between herbage accumu- The average behaviour of the farms showed that cow pregnancy rate
lation rate during and before the drought, was greater in paddocks and meat productivity decreased from 2007 to 2009 by 15% and
where high herbage heights were maintained (HHA), exhibiting 11%, respectively, and recovered from 2009 to 2010 (Figure 6). Farm
productivity values that ranged between 60% and 70% of the av- net income decreased from 2007 to 2008 and increased from 2008
erage productivity in the seasons before the drought. In contrast, to 2010.

TA B L E 3   Average values, standard

Variable HHA LHA Standard Error p‐Value
errors and P‐values (ANOVA) of herbage
height, herbage accumulation, stocking Herbage height (cm) 5.5 3.5 0.2 <.05
rate and animal live weight for contrasting Herbage accumulation (kg DM ha −1 −1
day ) 15.0 12.5 1.1 <.05
herbage allowance treatments in a 3 years
Stocking rate (kg LW ha−1 year−1) 382 398 7.0 NS
experiment in eastern Uruguay reported
by Do Carmo et al. (2018) Animal live weight (kg) 453 425 3.9 <.05
−1
Abbreviations: HHA, high herbage allowance (5.0 kg dry matter [DM] kg live weight [LW]); LHA,
low herbage allowance (3.5 kg dry matter [DM] kg−1 live weight [LW]).
 13652494, 2019, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/gfs.12445 by University Of Waterloo Dana Porter Library, Wiley Online Library on [02/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
|
642       MODERNEL et al.

TA B L E 4   Average values and p‐values (ANOVA) of resistance

and recovery of herbage accumulation rate and animal live
weight for contrasting herbage allowance treatments in a 3 years
experiment in eastern Uruguay reported by Do Carmo et al. (2018)

Herbage accumulation
rate Animal live weight

  HHA LHA p HHA LHA p

Resistancea 0.66 0.46 .03 0.98 0.94 <.01

b
Recovery 3.08 3.65 NS 20.8 18.1 NS

Abbreviations: HHA, high herbage allowance; LHA, low herbage

allowance.
a
Resistance is a ratio of values during drought and before drought with
same units, therefore unitless.
b
Recovery is calculated as difference between after drought and during
drought response variable over time (season). Units: kg DM ha day−1
season−1; and kg animal live weight season−1. F I G U R E 5   Resistance of herbage accumulation rate to drought
versus herbage height one month before the drought. Closed
circles indicate paddocks managed at high herbage allowance
Interestingly, the year when pregnancy rate and productivity (HHA), and open circles indicate paddocks managed at low herbage
reached a minimum (2009) was different from the year when farm allowance (LHA). The regression is significant at p < .01
income was minimum (2008). This finding may be understood, at
least in part, as a data‐reporting artefact, because the farm‐economy (Table S1) and negatively correlated with sheep‐to‐cattle ratio
variables are reported in fiscal years (from July 1st to June 30th). (r = −.81, p = .01) (Figure 7). These correlations suggest that farms
The drought period lasted from October 2008 to February 2009, with average high incomes and a low proportion of sheep in the herd
corresponding to the fiscal year 2008. The pregnancy rate after the withstand the drought better in terms of pregnancy rate.
drought was measured in the calving season (September–November) In order to study the “highest” level in our analytic hierarchy, the
2009 and therefore reported in the fiscal year 2009. The meat pro- farm, the nine farms were classified into four groups according to
ductivity affected was therefore the one from 2009. However, the two criteria: region (North Basalt vs. Eastern Sierras, Figure 3) and
higher expenses (costs) to withstand drought occurred during the resistance of pregnancy rate and productivity to drought (low vs.
fiscal year 2008. high). The structural variables of the farms were then analysed to
Cow pregnancy rate was affected by the drought because lower identify different strategies that might explain the differences.
herbage growth and supply caused weight loss of the animals. The Group 1 comprises farms located in the north, with a low pro-
drought particularly affected cow–calf systems since it occurred portion of sheep in the herd, high resistance of pregnancy rate and
during the mating season (December–February), when the nutri- meat productivity to drought, and high average pregnancy rates
tional status of the cows strongly determines the pregnancy rate (Do (farms North 1 and North 2). Group 2 comprises farms from the north
Carmo et al., 2016; Soca & Orcasberro, 1992). Farm meat productiv- (North 3, North 4 and North 5) with low resistance of pregnancy rate
ity is a direct function of pregnancy rate, since the most important to drought and high proportion of sheep in the herd. Group 3 com-
product of these farms are calves, and lower pregnancy rates mean prises farms from the Eastern Sierras, with high resistance of preg-
fewer calves, less livestock to market and lower meat productivity nancy rate to drought, but not of meat productivity and high levels
(Paparamborda, 2017). of improved pastures (East 1 and East 2). Finally, Group 4 comprises
farms (East 3 and East 4) which present high levels of resistance of
pregnancy rate and meat productivity to drought. These farms pres-
3.2.2 | Drought resistance and recovery
ent moderate levels of improved pastures and low sheep‐to‐cattle
While farm income is one of the most relevant variables required to ratios.
understand the sustainability of farming systems, it was not consid-
ered in the present investigation of resistance and recovery because Group 1 Extensive cattle farmers from the north. (North 1 and North 2)
many confounding variables can affect the report of farm income. As These farms show relatively high resistance of cow pregnancy and
a result, the resistance and recovery analysis were conceptualized meat productivity to drought. They share the same geographic loca-
using indices of cow pregnancy rate and meat productivity. These tion, similar low stocking rates, high meat productivity, a moderate
variables may be understood as both an independent (at the begin- sheep‐to‐cattle ratio and moderate recovery of meat productivity
ning of the drought period) and dependent variables (at the end of from drought. Both farms are among those with the highest incomes
the drought) in our conceptualization of this process. and the lowest input‐to‐output ratios. This group demonstrates that
Resistance of cow pregnancy rate was positively correlated with it is possible to achieve high productivity levels, resistance, recovery
pregnancy rate (r = .72, p = .02) and farm net income (r = .78, p = .02) and income simultaneously. Its pregnancy rate was 81% on average,
 13652494, 2019, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/gfs.12445 by University Of Waterloo Dana Porter Library, Wiley Online Library on [02/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
MODERNEL et al. |
      643

F I G U R E 7   Sheep‐to‐cow ratio and cow pregnancy rate for nine

farms in the Basalt and Eastern Sierras regions of Uruguay. The
straight line and equation indicate the linear regression between
variables. The regression is significant at p < .01

on shallow basaltic soils of the north of the country and usually

shows low levels of productivity and income (Paparamborda,
2017). This farm type was more common in the 1990s, but most
of the farms converted from sheep to cattle production after the
global drop in wool prices (Waquil, 2013). A common strategy
among thee farms, during droughts, is to accumulate sheep, be-
cause they harvest forage at lower herbage heights and are less
affected by drought than cattle (Nunes Gonçalves, 2007). These
farms display high input‐to‐output ratios, even though they usu-
ally have relatively low production costs. The low output in terms
of productivity reduces farm net income (Table 5). Farm North 3,
for example, displays pregnancy rates that decrease, down from
F I G U R E 6   Pregnancy rate (a), meat productivity (b) and farm levels around 80% to 50% in 2009 (Figure 8c). Those low rates
net income (c) for the nine livestock grazing farms during 4 years did not recover after the drought. Productivity levels decrease to-
in Uruguay. Vertical lines indicate standard deviations. The grey wards 2009 but recover afterwards (Figure 8c). The trend in farm
shadow area indicates the drought period net income is similar to that of meat productivity, with lower val-
ues in 2008 than in 2007 and recovery afterwards (Figure 8d). The
and Figure 8a,b shows the behaviour of North 1 over time. The fig- input‐to‐output ratio increases from 2007 to 2008. While meat
ure shows that meat productivity decreased in 2008 but increased productivity levels were relatively high (110–130 kg LW/ha), these
sharply in 2009 (Figure 8b). While farm net income was low dur- did not translate into higher income. This may be explained by the
ing 2008 and 2010, the group's low input‐to‐output ratio (Table 4) fact that 50% of the productivity of the farm consisted of sheep
maintains farm income at reasonable levels compared to the entire meat (30%) and wool (20%), which have considerably lower market
population of farms. prices than beef.

Group 2 Sheep-oriented farmers from the north Group 3 Intensive cattle farmers on improved pastures
These farms (North 3, North 4 and North 5) show the lowest re- These farms have high resistance of cow pregnancy but low resistance
sistance (through cow pregnancy rates) and variable resistance of meat productivity to drought. For instance, East 1 had the highest
of meat productivity. This group of farms represents traditional resistance of pregnancy rate (1.16), which means that, in contrast with
cow–calf farms, coupled with a high proportion of sheep in the the other farms, pregnancy rate increased during the drought (Table 4).
herd, and a low percentage of improved pastures (Table 4). Farms While pregnancy rate was not affected during the drought year, pos-
belonging to this group have high stocking rates (hence low herb- sibly because the farm sold cows with low body condition, or supple-
age allowances), low cow pregnancy rates, the lowest incomes, mented grazing with external feeds (Figure 8e), meat productivity was
the lowest resistance of pregnancy rate and variable resistance of the lowest. This type of farm represents the highest input‐to‐output
meat productivity. This type of farm configuration is very common ratio within our sample, which means that, on average over the 3 years,
 13652494, 2019, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/gfs.12445 by University Of Waterloo Dana Porter Library, Wiley Online Library on [02/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
|
644       MODERNEL et al.

F I G U R E 8   Trajectories of cow pregnancy rate (%), meat productivity (kg LW/ha), input/output ratio and farm net income (kg US$ ha−1)
for farms North 1 (a and b), North 3 (c and d), East 1 (e and f) and East 3 (g and h) during the period 2007–2010. Panels on the left indicate
meat productivity (full lines) and pregnancy rate (dashed lines). Panels on the right indicate economic input–output ratio (full lines) and farm
net income (dashed lines). Resistance indicators are identified for each farm as Rp (resistance pregnancy rate) and Rm (resistance meat
productivity)

this type spent 29% more of its gross farm income than the other farm is to increase productivity and farm income through higher inputs (ex-
types (Table 4). While farm income is, on average, positive, the large pressed by the fact that 23% of the grazing area is under sown pasture
fluctuations between years suggest high vulnerability to external as opposed to native grassland). The high investment needed for im-
shocks. Deeper investigation showed that this type of farm followed proving the herbage quantity and quality in the long term increases
a conventional intensification pathway proposed for the region, that the financial risk of the farm, and the exotic grass and legume species
 13652494, 2019, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/gfs.12445 by University Of Waterloo Dana Porter Library, Wiley Online Library on [02/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
MODERNEL et al. |
      645

TA B L E 5   Structural and management variables, resistance and recovery for each farm of the study. Structural and management variables
calculated as averages for the period 2007–2010

  North 1 North 2 North 3 North 4 North 5 East 1 East 2 East 3 East 4

Group 1 1 2 2 2 3 3 4 4
Soil productivity indexa 63 109 101 64 120 88 270 58 58
Grazing area (ha) 1100 696 157 990 398 354 247 904 743
Area improved pastures (%) 0.0 8.0 4.0 6.0 5.4 23 22 12 12
Stocking rate (LU/ha) 0.6 0.6 1.2 0.7 0.9 0.7 0.9 0.8 0.8
Sheep‐to‐cattle ratiob 2.1 1.3 4.7 5.7 7.6 1.6 2.7 0.1 0.3
Cow pregnancy rate (%)b 81 81 65 69 64 81 84 69 69
Meat productivity (kg/ha) 84 89 121 73 90 82 109 71 65
Farm net income (US$/ha) 35 17 4.0 8.0 ‐8.0 19 31 22 16
Input‐to‐output ratiob 0.44 0.61 1.06 0.81 0.93 1.29 0.72 0.36 0.72
Resistance of pregnancy rate 0.95 0.88 0.63 0.69 0.58 1.16 0.91 0.89 0.96
Resistance of meat 0.93 0.90 0.88 0.86 1.26 0.48 0.81 1.60 0.99
productivity
Recovery of pregnancy rate ‐7.0 22 0.0 15 28 ‐2 ‐11 16 26
Recovery of meat 7.0 2.0 20 9.0 ‐7.0 ‐24 21 14 31
productivity

Abbreviation: LU, livestock unit.

a
‘CONEAT’ index, used in Uruguay as a proxy for soil quality (MGAP, 2018). The average index for the country is 100.
b
Ratio between costs (input) and income (product).

sown may be less adapted to recurrent drought and less resistant to accumulation and animal weight to drought. Second, at the farm
these events, risking productivity. While this intensification pathway level, managing (and minimizing) the sheep‐to‐cow ratio is highly
has been promoted as “sustainable intensification” from an economic related to increasing resistance of cow pregnancy rate to drought.
and environmental perspective (Dick, Abreu da Silva, & Dewes, 2015; The correlations are valid for the nine farms that represent four
de Oliveira Silva et al., 2016; Pashaei et al., 2016), recent studies show “idealized” farm types in Uruguay. They suggest interesting systemic
that higher productivity does not necessarily translate in higher in- relations. We would suggest caution in the acceptance of the find-
come, and can even worsen environmental performance as measured ings; the more interesting findings must be further investigated and
by biodiversity, fossil fuel energy consumption and nutrient balances cannot be generalized without further evidence.
(Modernel et al., 2018). Nevertheless, this analysis at the farm level constitutes, as far as
the authors know, the first drought‐related resilience assessment of
Group 4 Extensive cattle farmers of the east native grassland‐based farming systems. Four farm production strat-
These farms (East 3 and East 4) have high resistance of cow preg- egies were identified that result in different responses to drought.
nancy and meat productivity to drought. Both farms show similarities These include: implementing low stocking rates (Group 1), maintain-
with the farmers of the north given the highest levels of resistance ing high proportions of sheep in the herd (Group 2), increasing the
of pregnancy rate and productivity and high levels of farm income. amount of herbage biomass through improved pastures (Group 3)
On the other hand, they differ in structural variables. This type of and keeping only cattle (Group 4).
farms has lower sheep‐to‐cattle ratios, higher stocking rate and Group 1 and Group 4 farms were the most resistant in terms of
higher area of improved pastures than Group 1 farms. Interestingly, cow pregnancy rate and meat productivity. Maintaining a low sheep‐
the pregnancy rates and productivity levels are among the lowest of to‐cattle ratio promotes higher cow pregnancy rates and translates
the entire population of farms. Farm East 3 illustrates this. The preg- into higher productivity. Future research might investigate if/how
nancy rate decreased from 2007 to 2008 and continued at low lev- higher herbage biomass levels result from lower stocking rates and
els in 2009. Meat productivity was strongly affected by the drought from less competition between sheep and cows for grazing.
(63% decrease) but recovered, although at lower levels than before Group 2 and Group 3 farms seem systemically unbalanced. Group
the drought (Figure 8). 2 by dominance of sheep in the herd (a livestock species that can
better resist droughts than cattle but has a low current economic
return). Group 3 by investing in highly “productive” pastures that,
3.3 | Integration of paddock and farm level
in the short term, can lead to increased forage productivity and in-
Two main findings arise from this study. First, managing herbage creased animal productivity. But the non‐native species employed
allowance at the paddock level increases resistance of herbage to improve pastures are not well adapted to drought, increase the
 13652494, 2019, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/gfs.12445 by University Of Waterloo Dana Porter Library, Wiley Online Library on [02/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
|
646       MODERNEL et al.

ecological value of sheep as a species. This includes increasing the

diversification of products (lamb and wool) and the ability to con-
sume non‐desired grass species by cattle.

4 | CO N C LU S I O N S

Grazing management strategies have an important impact on abil-

ity of grazing systems to resist to droughts. Higher pre‐drought
herbage height increases resistance of herbage accumulation and
animal weight, while lower sheep‐to‐cow ratios increase the resist-
ance of cow pregnancy rate (a key variable for the income of cow–
calf farmers in the region). Unexpectedly, no differences between
grazing regimes were found for recovery of herbage accumulation
rate or animal weight on the native grasslands. This suggests that
regardless of the grazing management, grazing systems on native
F I G U R E 9   Management decisions at paddock and farm level, grasslands can recover quickly when rainfall resumes after drought.
the key variables they are directly impacting and the outcomes for This result must be taken with caution since overgrazing situations
the farm performance and resilience
might be more severe than those created by the low herbage al-
lowance treatment of the experiment. The results of this study can
vulnerability of the local grazing system to drought and incur greater inform farmers and policymakers to formulate strategies to mitigate
economic cost and risk. Future research in these areas might investi- the frequently occurring droughts in the region. These strategies
gate the relative returns to investment in cattle and sheep in drought should be based on grazing management which demand low‐cost
areas, as well as the relative returns to pasture improvement under technologies and can prevent extremely negative impacts on pro-
drought conditions in grassland Uruguay. duction systems.
The results of this study may aid in the redesign of sustainable and
more resilient mixed livestock farming systems of the grassland re-
AC K N OW L E D G M E N T S
gions of Uruguay. Figure 9 represents the relation between the study
levels. It synthesizes our main findings at farm and paddock level, hy- This work has been conducted as part of a PhD thesis project of
pothesizing that farms that perform better in cow pregnancy rate and P. Modernel supported by the Agricultural Transformation by
farm income are a result of higher forage height at paddock level. Innovation (AGTRAIN) Erasmus Mundus Joint Doctorate Program,
Further research should link these two levels, either through funded by the EACEA (Education, Audiovisual and Culture Executive
farm surveys or on‐station experiments. By integrating these two Agency) of the European Commission. Partial funding was received
levels, it should be underlined that sheep alone are not “the prob- from CSIC—UDELAR grant to V. Picasso (Sustentabilidad de los sis-
lem”, but rather that low forage allowance and high sheep‐to‐cow temas de producción agropecuarios). We thank Sarah Bullock and
ratios seem directly related to low levels of farm productivity under Todd Zdorkowski for helpful suggestions to the final version of the
drought conditions. This is expressed by forage allowance at paddock manuscript.
level and not by stocking rate, as was previously found by Do Carmo
et al. (2018), Do Carmo et al. (2016) and discussed by Sollenberger
ORCID
et al. (2005). While stocking rate is a relevant metric for explaining
farm performance, forage allowance (kg grass dry matter per kg ani- Pablo Modernel  https://orcid.org/0000-0002-2858-009X
mal body weight) is the key variable for grassland management. This Valentin Picasso  https://orcid.org/0000-0002-4989-6317
means that stocking rate ought to be a function of forage allowance Martin Do Carmo  https://orcid.org/0000-0001-7632-8449
rather than an independent variable. The results on this topic found
in this study support previous findings by Ruggia et al. (2015) and
Scarlato et al. (2015). Those authors showed that the most important REFERENCES

factor in the redesign of mixed livestock farming systems, resulting in ’T Mannetje, L. (1978). Measurement of grassland vegetation and animal
improved pregnancy rates and farm incomes, is a decreasing sheep‐ production. Hurley, Berkshire: Commonwealth Agricultural Bureaux,
to‐cattle ratio. This strategy reduces competition among sheep and Hurley.
Ahmed, A. G. M., Azeze, A., Babiker, M., & Tsegaye, D. (2002). Post‐
cattle for grazing, favours the energy balance of cattle, improves cow
drought recovery strategies among the pastoral households in the Horn
pregnancy rates and, thereby, increases farm incomes. of Africa: A review. Dev. Res. Rep. Ser. 80.
Even though the role of sheep seems detrimental in productivity Allen, V. G., Batello, C., Berretta, E. J., Hodgson, J., Kothmann, M., Li,
levels and farm income, it does not undermine the economic and X., … Sanderson, M. (2011). An international terminology for grazing
 13652494, 2019, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/gfs.12445 by University Of Waterloo Dana Porter Library, Wiley Online Library on [02/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
MODERNEL et al. |
      647

lands and grazing animals. Grass & Forage Science, 66, 2–28. https​:// of Agricultural Sustainability, 8, 186–198. https​://doi.org/10.3763/
doi.org/10.1111/j.1365-2494.2010.00780.x ijas.2010.0480
Barros, V., Clarke, R., & Dias, P. S. (2006). El Cambio Climático En La Cuenca de Oliveira Silva, R., Barioni, L. G., Hall, J. A. J., Folegatti Matsuura, M.,
Del Plata. Buenos Aires. Zanett Albertini, T., Fernandes, F. A., & Moran, D. (2016). Increasing
Bartaburu, D., Duarte, E., Montes, E., Morales Grosskopf, H., & Pereira, beef production could lower greenhouse gas emissions in Brazil if
M. (2009). Las sequías : un evento que afecta la trayectoria de las em- decoupled from deforestation. Nature Climate Change, 6, 3–8. https​
presas y su gente. In H. Morales Grosskopf, & F. Dieguez Cameroni ://doi.org/10.1038/nclim​ate2916
(Eds.), Familias y Campo: Rescatando Estrategias de Adaptación. Dick, M., Abreu da Silva, M., & Dewes, H. (2015). Mitigation of envi-
Instituto Plan Agropecuario (pp. 155–168). Montevideo: Instituto Plan ronmental impacts of beef cattle production in southern Brazil e
Agropecuario. Evaluation using farm‐based life cycle assessment. Journal of Cleaner
Becoña, G., Astigarraga, L., & Picasso, V. D. (2014). Greenhouse gas Production, 87, 58–67. https​://doi.org/10.1016/j.jclep​ro.2014.01.080
emissions of beef cow‐calf grazing systems in Uruguay. Sustainable Do Carmo, M., Claramunt, M., Carriquiry, M., & Soca, P. (2016). Animal
Agriculture Research, 3, 89. https​://doi.org/10.5539/sar.v3n2p89 energetics in extensive grazing systems: Rationality and results of re-
Berretta, E. J., Risso, D. F., Montossi, F., & Pigurina, G. (2000). Campos in search models to improve energy efficiency of beef cow‐calf grazing
Uruguay, Grassland ecophysiology and grazing ecology. Cambridge: CABI. Campos systems. Journal of Animal Science, 94, 84–92. https​://doi.
Bestelmeyer, B. T., & Briske, D. D. (2012). Grand challenges for resilience‐ org/10.2527/jas20​16-0596
based management of rangelands. Rangeland Ecology & Management, Do Carmo, M., Sollenberger, L. E., Carriquiry, M., & Soca, P. (2018).
65, 654–663. https​://doi.org/10.2111/REM-D-12-00072.1 Controlling herbage allowance and selection of cow genotype im-
Bilenca, D., & Miñarro, F. (2004). Identificación de Áreas Valiosas de prove cow‐calf productivity in Campos grasslands. Professional Animal
Pastizal (AVPs) en las Pampas y Campos de Argentina, Uruguay y sur de Scientist, 34, 32–41. https​://doi.org/10.15232/​pas.2016-01600​
Brasil. Buenos Aires: Fundación Vida Silvestre Argentina. Douxchamps, S., Van Wijk, M. T., Silvestri, S., Moussa, A. S., Quiros, C.,
Bresciano, D., Rodrigues, C., Lezama, F., & Altesor, A. (2014). Patrones Ndour, N. Y. B., … Rufino, M. C. (2016). Linking agricultural adapta-
de invasión de los pastizales de Uruguay a escala regional. Ecología tion strategies, food security and vulnerability: Evidence from West
Austral, 24, 83–93. Africa. Regional Environmental Change, 16, 1305–1317. https​://doi.
Briske, D. D., Derner, J. D., Brown, J. R., Fuhlendorf, S. D., Teague, W. R., org/10.1007/s10113-015-0838-6
Havstad, K. M., … Willms, W. D. (2008). Rotational grazing on rangelands: FAO (2013). Clima de cambios. Nuevos desafíos de adaptación en Uruguay.
Reconciliation of perception and experimental evidence. Rangeland Montevideo, Uruguay: Food and agriculture organization.
Ecology & Management, 61, 3–17. https​://doi.org/10.2111/06-159R.1 Focht, T., & Borges de Medeiros, R. (2012). Prevention of natural grass-
Cabel, J. F., & Oelofse, M. (2012). An indicator framework for assessing land invasion by Eragrostis plana Nees using ecological management
agroecosystem resilience. Ecology and Society, 17(1), 18. https​://doi. practices. Revista Brasileira De Zootecnia, 41, 1816–1823. https​://doi.
org/10.5751/ES-04666-170118 org/10.1590/S1516-35982​01200​0 800003
Caffera, R. M. (2005). Variación de largo periodo en la disponibilidad poten‐ Godfray, H. C. J., Beddington, J. R., Crute, I. R., Haddad, L., Lawrence, D.,
cial de agua para pasturas en Uruguay. Buenos Aires: Universidad de Muir, J. F., … Toulmin, C. (2010). Food security: The challenge of feed-
Buenos Aires ing 9 billion people. Science, 327, 812–818. https​://doi.org/10.1126/
Carvalho, P. C., & Batello, C. (2009). Access to land, livestock production scien​ce.1185383
and ecosystem conservation in the Brazilian Campos biome: The nat- Grimm, V., & Wissel, C. (1997). Babel, or the ecological stability discussions:
ural grasslands dilema. Livestock Science, 120, 158–162. https​://doi. An inventory and analysis of terminology and a guide for avoiding confu-
org/10.1016/j.livsci.2008.04.012 sion. Oecologia, 109, 323–334. https​://doi.org/10.1007/s0044​20050090
Carvalho, P. C. F., Soares, A. B., Garcia, É. N., Boldrini, I. I., Pontes, L. S., Groot, J., Cortez‐Arriola, J., Rossing, W., Améndola Massiotti, R., &
Velleda, G. L., …Júnior, J. A. F. (2003). Herbage allowance and species Tittonell, P. (2016). Capturing agroecosystem vulnerability and resil-
diversity in native pastures, In: Proceedings of the VIIth International ience. Sustainability, 8(11), 1206. https​://doi.org/10.3390/su811​1206
Rangelands Congress. Durban, pp. 858–859. Gutierrez, R., & Modernel, P. (2011). Los procesos de adopción y manejo
Claramunt, M., Fernández‐Foren, A., & Soca, P. (2017). Effect of herbage tecnológico en la producción familiar criadora. Revista Del Plan
allowance on productive and reproductive responses of primiparous Agropecuario, 140, 60–63.
beef cows grazing on Campos grassland. Animal Production Science, Haydock, K., & Shaw, N. (1975). The comparative yield method for esti-
58(9), 1615. mating dry matter yield of pasture. Australian Journal of Experimental
Cobon, D. H., Stone, G. S., Carter, J. O., Scanlan, J. C., Toombs, N. R., Agriculture, 15, 663. https​://doi.org/10.1071/EA975​0663
Zhang, X., … McKeon, G. M. (2009). The climate change risk man- Holling, C. S. (1973). Resilience and stability of ecological systems. Annual
agement matrix for the grazing industry of northern Australia. The review of ecology and systematics, 4(1), 1–23.
Rangeland Journal, 31, 31–49. https​://doi.org/10.1071/RJ08069 Holling, C. S. (1996). Engineering resilience versus ecological resilience.
Cruz, G., Baethgen, W., Bartaburu, D., Bidegain, M., Giménez, A., In P. C. Schulze (Ed.), Engineering within ecological constraints (pp.
Methol, M., … Vinocur, M. (2018). Thirty years of multilevel pro- 31–44). Washington, DC: National Academy of Engineering National
cesses for adaptation of livestock production to droughts in Uruguay. Academy Press. https​://doi.org/10.17226/​4919.
Weather, Climate, and Society, 10, 59–74. https​://doi.org/10.1175/ INUMET (2018). Estadísticas climatológicas [WWW Document]. https​://
WCAS-D-16-0133.1 www.inumet.gub.uy/clima/​estad​istic​as-clima​tolog​icas (Accessed
Cruz, G., Baethgen, W., Picasso, V., & Terra, R. (2014). Análisis de sequías July 26, 2017) .
agronómicas en dos regiones ganaderas de Uruguay. Agrociencia IPCC (2013). Fifth Assessment Report – Climate Change 2013 [WWW
Uruguay, 18, 126–132. Document]. Retrieved from http://www.ipcc.ch/repor​t /ar5/wg1/
Da Trindade, J. K. J. K., Pinto, C. E., Neves, F. P., Mezzalira, J. C., Bremm, Ives, A. R., & Carpenter, S. R. (2007). Stability and diversity of ecosys-
C., Genro, T. C. M., … Carvalho, P. C. F. (2012). Forage allowance as a tems. Science, 317, 58–62. https​://doi.org/10.1126/scien​ce.1133258
target of grazing management: Implications on grazing time and for- López‐Ridaura, S., Keulen, H. V., Ittersum, M. K. V., Leffelaar, P. A.
age searching. Rangeland Ecology & Management, 65, 382–393. https​ (2005a). Multiscale methodological framework to derive criteria and
://doi.org/10.2111/rem-d-11-00204.1 indicators for sustainability evaluation of peasant natural resource
Darnhofer, I., Fairweather, J., & Moller, H. (2010). Assessing a farm’s management systems. Environment, Development and Sustainability,
sustainability: Insights from resilience thinking. International Journal 7, 51–69. https​://doi.org/10.1007/s10668-003-6976-x
 13652494, 2019, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/gfs.12445 by University Of Waterloo Dana Porter Library, Wiley Online Library on [02/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
|
648       MODERNEL et al.

López‐Ridaura, S., van Keulen, H., van Ittersum, M. K., & Leffelaar, P. A. Picasso, V. D., Casler, M. D., & Undersander, D. (2019). Resilience, stabil-
(2005b). Multi‐scale sustainability evaluation of natural resource ity, and productivity of Alfalfa cultivars in rainfed regions of North
management systems: Quantifying indicators for different scales of America. Crop Science. https​://doi.org/10.2135/crops​ci2018.06.0372
analysis and their trade‐offs using linear programming. International R Core Team (2015). R: A language and environment for statistical com-
Journal of Sustainable Development & World Ecology, 12, 81–97. https​:// puting. Boston, MA: R Core Team.
doi.org/10.1080/13504​50050​9469621 Risso, D., Ayala, W., Bermúdez, R., & Berretta, E. (2005). Seminario de
Maraschin, G. E. (2001). Production potential of South America grasslands. actualización técnica en manejo de campo natural, Serie Técn. ed, Série
XIX Int. Grassl. Congr. Brazil, Proc. 5–15. técnica 151. INIA, Montevideo, Uruguay.
Marengo, J. A., Liebmann, B., Grimm, A. M., Misra, V., Silva Dias, P. L., Roesch, L. F., Vieira, F., Pereira, V., Schünemann, A. L., Teixeira, I., Senna,
Cavalcanti, I. F. A., … Alves, L. M. (2012). Recent developments A. J., & Stefenon, V. M. (2009). The Brazilian pampa: A fragile biome.
on the South American monsoon system. International Journal of Diversity, 1, 182–198. https​://doi.org/10.3390/d1020182
Climatology, 32, 1–21. https​://doi.org/10.1002/joc.2254 Ruggia, A., Scarlato, S., Cardozo, G., Aguerre, V., Dogliotti, S., Rossing,
MGAP (2015). DIEA. Anuario estadístico agropecuario. Minist. Ganad. W., & Tittonell, P. (2015). Managing pasture‐herd interactions in live-
Agric. y pesca 215 stock family farm systems based on natural grasslands in Uruguay.
MGAP (2018). Descripción de Grupos de Suelos CO.N.E.A.T. Montevideo. In G. Emmanuel, & J. Wery (Eds.), 5th International Symposium
[WWW Document]. http://www.mgap.gub.uy/sites/​defau​lt/files/​ for Farming Systems Design “Multi‐Functional Farming Systems in a
descr​ipcion_de_grupos_de_suelos_coneat.pdf Changing World” (pp. 267–268). Montpellier: France.
Modernel, P., Astigarraga, L., & Picasso, V. (2013). Global versus local Scarlato, S., Albicette, M. M., Bortagaray, I., Ruggia, A., Scarlato, M., &
environmental impacts of grazing and confined beef production Aguerre, V. (2015). Co‐innovation as an effective approach to pro-
systems. Environmental Research Letters, 8, 035052. https​://doi. mote changes in farm management in livestock systems in Uruguay.
org/10.1088/1748-9326/8/3/035052 In G. Emmanuel (Ed.), 5th International Symposium for Farming Systems
Modernel, P., Dogliotti, S., Alvarez, S., Corbeels, M., Picasso, V., Tittonell, Design “Multi‐Functional Farming Systems in a Changing World” (pp.
P., & Rossing, W. A. H. (2018). Identification of beef production farms 281–282). Montpellier: France.
in the Pampas and Campos area that stand out in economic and en- Shiu, C. J., Liu, S. C., Fu, C., Dai, A., & Sun, Y. (2012). How much do precip-
vironmental performance. Ecological Indicators, 89, 755–770. https​:// itation extremes change in a warming climate? Geophysical Research
doi.org/10.1016/j.ecoli​nd.2018.01.038 Letters, 39, 1–5. https​://doi.org/10.1029/2012G​L052762
Modernel, P., Rossing, W. A. H., Corbeels, M., Dogliotti, S., Picasso, Soares, A. B., Carvalho, P. C. F., Nabinger, C., Frizzo, A., Pinto, C. E., Júnior,
V., & Tittonell, P. (2016). Land use change and ecosystem service J. A. F., … Trindade, J. K. (2003). Effect of changing herbage allow-
provision in Pampas and Campos grasslands of southern South ance on primary and secondary production of natural pasture. In A. R.
America. Environmental Research Letters, 11, 113002. https​://doi. Allsopp, S. J. Palmer, K. P. Milton, G. I. H. Kirkman, C. R. Kerley, & C.
org/10.1088/1748-9326/11/11/113002 J. B. Hurt (Eds.), VIIth International Rangelands (pp. 966–968). Durban.
Norton, M. R., Malinowski, D. P., & Volaire, F. (2016). Plant drought Soca, P., Olmos, F., Espasandín, A., Bentancur, D., Pereyra, F., Cal, V., …
survival under climate change and strategies to improve perennial Do Carmo, M. (2008). Impacto de cambios en la estrategia de asig-
grasses. A review. Agronomy for Sustainable Development, 36, 29. nación de forraje sobre la productividad de la cría con diversos gru-
https​://doi.org/10.1007/s13593-016-0362-1 pos genéticos bajo pastoreo de campo natural. Semin. Actual. técnica
Nunes Gonçalves, E. (2007). Comportamiento ingestivo de bovinos e Cría Vacuna, 110–119.
ovinos em pastagem natural da depresao central do Rio Grande do Soca, P., & Orcasberro, R. (1992). Propuesta de Manejo del Rodeo de Cría
Sul. Universidade Federal do Rio Grande do Sul. en base a estado corporal, altura del pasto y aplicación del destete
Oficialedgui, R. (1985). Carne equivalente: Los riesgos de la simplificación. temporario. In Jornada de Producción Animal, Paysandú Evaluación
Rev. la AIA 272–279. Física y Económica de Alternativas Tecnológicas En Predios Ganaderos
Oliver, T. H., Heard, M. S., Isaac, N. J. B., Roy, D. B., Procter, D., Eigenbrod, (pp. 54–56). Montevideo: Facultad de Agronomía.
F., … Bullock, J. M. (2015). Biodiversity and resilience of ecosystem Sollenberger, L. E. (2015). Challenges, opportunities, and applications of
functions. Trends in Ecology & Evolution, 30, 673–684. https​://doi. grazing research. Crop Science, 55, 2540. https​://doi.org/10.2135/
org/10.1016/j.tree.2015.08.009 crops​ci2015.02.0070
Overbeck, G. E., Mueller, S. C., Fidelis, A., Pfadenhauer, J., Pillar, V. Sollenberger, L. E., Moore, J. E., Allen, V. G., & Pedreira, C. G. S. (2005).
D. D., Blanco, C. C., … Froneck, E. D. (2007). Brazil’s neglected Reporting forage allowance in grazing experiments. Crop Science, 48,
biome: The South Brazilian Campos. Perspectives in Plant Ecology, 896–900. https​://doi.org/10.2135/crops​ci2004.0216
Evolution and Systematics, 9, 101–116. https​://doi.org/10.1016/j. Soriano, A. (1992). Río de la Plata Grasslands. In R. T. Coupland (Ed.),
ppees.2007.07.005 Ecosystems of the world (pp. 367–407). Amsterdam, the Netherlands:
Overbeck, G. E., Müller, S. C., Pillar, V. D., & Pfadenhauer, J. (2006). Elsevier.
Floristic composition, environmental variation and species distribu- Teague, R., Provenza, F., Norton, B., & Steffens, T. (2009). Benefits of
tion patterns in burned grassland in southern Brazil. Brazilian Journal multi‐paddock grazing management on rangelands: Limitations of
of Biology, 66, 1073–1090. https​://doi.org/10.1590/S1519-69842​ experimental grazing research and knowledge gaps. In H. Schöder
00600​0600015 (Ed.), Grasslands: Ecology, management and restoration (pp. 1–40).
Paolino, C., Methol, M., & Quintans, D. (2010). Estimación del impacto de Hauppauge, NY: Nova publishers
una eventual sequía en la ganadería nacional y bases para el diseño de Thurow, T. L., & Taylor, C. A. (1999). Viewpoint: The role of drought in
políticas de seguros. Montevideo. range management. Journal of Range Management, 52, 413–419. https​
Paparamborda, I. (2017). ¿Qué nos dicen las prácticas de gestión del pas‐ ://doi.org/10.2307/4003766
toreo en los predios ganaderos familiares sobre su funcionamiento y re‐ Tittonell, P. (2014). Livelihood strategies, resilience and transformability
sultado productivo? Universidad de la República. in African agroecosystems. Agricultural Systems, 126, 3–14. https​://
Pashaei, F., van der Linden, A., Meuwissen, M. P. M., Cunha Malafaia, G., doi.org/10.1016/j.agsy.2013.10.010
Oude Lansink, A. G. J. M., & de Boer, I. J. M. (2016). Environmental Urruty, N., Tailliez‐Lefebvre, D., & Huyghe, C. (2016). Stability, robust-
and economic performance of beef farming systems with different ness, vulnerability and resilience of agricultural systems. A re-
feeding strategies in southern Brazil. Agricultural Systems, 146, 70– view. Agronomy for Sustainable Development, 36, 1–15. https​://doi.
79. https​://doi.org/10.1016/j.agsy.2016.04.003 org/10.1007/s13593-015-0347-5
 13652494, 2019, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/gfs.12445 by University Of Waterloo Dana Porter Library, Wiley Online Library on [02/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
MODERNEL et al. |
      649

van Apeldoorn, D. F., Sonneveld, M. P. W., & Kok, K. (2011). Landscape systems. Ecology and Society, 9, 9. https​://doi.org/10.5751/
asymmetry of soil organic matter as a source of agro‐ecosystem re- ES-00650-090205
silience. Agriculture, Ecosystems & Environment, 140, 401–410. https​ Waquil, P. D. (2013) The evolution of sheep production in Rio Grande do Sul
://doi.org/10.1016/j.agee.2011.01.002 and Uruguay: a comparative analysis of structural change 1134–1140.
Van Ruijven, J., & Berendse, F. (2010). Diversity enhances community
recovery, but not resistance, after drought. Journal of Ecology, 98,
81–86. https​://doi.org/10.1111/j.1365-2745.2009.01603.x
How to cite this article: Modernel P, Picasso V, Do Carmo M,
Vogel, A., Scherer‐Lorenzen, M., & Weigelt, A. (2012). Grassland resis-
et al. Grazing management for more resilient mixed livestock
tance and resilience after drought depends on management intensity
and species richness. PLoS ONE, 7, e36992. https​://doi.org/10.1371/ farming systems on native grasslands of southern South
journ​al.pone.0036992 America. Grass Forage Sci. 2019;74:636–649. https​://doi.
Walker, B., Holling, C. S., Carpenter, S. R., & Kinzig, A. (2004). org/10.1111/gfs.12445​
Resilience, adaptability and transformability in social–ecological