Growing season aridity was calculated using a long-term dataset
collected at the CDRRC from 1967-2004 (as described by Beck et
al. 2007). We analyzed the De Martonne Aridity Index for the mean
monthly precipitation (mm) and temperature (°C) over the four months
of June, July, August, and September (Equation 2). For a given year,
the Growing Season Aridity Index was calculated by taking the mean
monthly precipitation (June through September, divided by four) and
dividing by the mean monthly temperature (June through September,
divided by four) plus 10. Similar to the annual De Martonne aridity
Index, when presented, high Growing Season De Martonne aridity (GS
IDM) scores indicate cooler and wetter growing seasons, and lower
aridity scores indicate hotter and drier conditions. Thus, high GS
IDM scores demonstrate low growing season aridity, and low GS IDM
scores high growing season aridity.
For this study, cattle were rotated through four pastures in
1967-2004 (Beck et al., 2007). Cattle were rotated through the
pastures depending on four assigned seasonality treatments: 1) a
yearlong grazing pasture (1267 ha), 2) a winter-spring grazing
pasture (508 ha), 3) a fall grazing pasture (670 ha), and 4) a
summer grazing pasture (494 ha; Fig. 2). The yearlong pasture was
grazed continuously every year. The summer pasture was grazed from
June to the middle of September. The fall pasture was grazed from
the middle of September to the end of December. The winter/spring
pasture was grazed from January to late June (Fig. 2). To keep
stocking rates similar across the grazing season treatments, the
number of cows in a pasture for a 3-month grazing season was 4-fold
greater than the number of cows in the yearlong pasture.
Cattle used in the study from 1967 to 1971 were Hereford; Brangus
cattle herds were used from 1972 to 1992. A mixed herd (different
breeds) was used from 1992 through 2004, and no cattle grazed the
pastures in 1995 and 1996 due to drought and poor forage conditions.
At the beginning of the study, stocking rates were established at a
conservative rate and herd sizes were monitored to adjust to
changing conditions. Stocking rate increased in the 1980’s when
perennial productivity increased due to increased rainfall. Each
pasture was grazed conservatively to attempt to meet similar
utilization rates on the dominant perennial grasses for the
respective year and stocking density was standardized across
pastures to account for pasture size differences. If the number of
cattle in any pasture was adjusted, due to forage limitations or
other phenomenon (e.g. sick cows needing to be removed), the number
of cattle was also adjusted in the other pastures to maintain a
standard number of animal units. Cattle stocking rates were based on
previous year’s perennial grass production and maintaining a
bull-cow ratio.
In each pasture and grazing treatment, the west end of an individual
transect was randomly located and then laid out due east from the
starting point. The original study included 220 total transects,
however for this study, a total of 78, 61m (200 ft) permanent
transects were established and scaled to the size of the pasture: 35
transects in the yearlong pasture, 20 transects in the winter-spring
pasture, 12 transects in the summer pasture, and 11 transects in the
fall pasture (Fig. 2). The yearlong and fall pastures shared a fence
line water source (32°34’46.3”N 106°55’19.1”W) as did the
winter-spring and summer pastures (32°35’11.8”N 106°52’23.0”W; Fig.
2). To reduce the confounding influence of distance from water, we
focused only on transects within 1609.34 meters (1 mile; the point
at which livestock grazing use begins to diminish) of drinkers thus
reducing transect totals from 220 to 78 in total. For comparison
purposes across vegetation and soil characteristics, the four
pastures, and associated transects, fell within similar ecological
sites and states.
Perennial grass productivity (kg/ha) was measured annually at the
end of each growing season in mid-September through mid-October,
1967-2002 and 2004. To directly measure current year’s production,
our three target species (black grama, threeawns, and dropseed that
were taller than 2 cm) were clipped from five 0.3-m2 plots
distributed every 12 m along each transect. To account for previous
year’s impact of sampling, the plots were moved 1 meter in either
direction from year to year. The clipped biomass was then dried for
seventy-two hours at 66°C. After the biomass samples had been dried,
they were weighed and averaged at the transect level. Accordingly,
our biomass measurements are reported as kilograms of dry matter
(DM) per hectare (kg DM*ha-1). Following these protocols for the
duration of the 37-year study, vegetation baseline conditions, as
well as their variability over time were able to be assessed for
total biomass (Fig 3 a) and our individual grass taxa (Fig 3 b-d).
Temperature data used for this study were derived from the following
data set:
Wooton, E., National Weather Service, D.
Thatcher, J. Anderson, and K. Havstad. 2022. Locally verified daily
temperature and precipitation data from a NOAA weather station at
USDA Jornada Experimental Range headquarters, southern New Mexico
USA, 1914-2006 ver 81. Environmental Data Initiative.
https://portal.edirepository.org/nis/mapbrowse?scope=knb-lter-jrn&identifier=210379001
The average annual temperature was calculated in the following
manner. First, we calculated the daily mean temperature by adding
the daily minimum temperature and daily maximum temperature and
dividing by two. We then averaged the daily mean values to find the
monthly average temperature for each month. The mean of the twelve
monthly averages provided the average annual temperature.
Precipitation data used for this study are included as a data entity
in this data package.
