Abstract
In the last 5 years, an average of 7.7 million acres have been burned annually by wildfires inthe United States, with the western states being particularly prone. Previous studies have found that
wildfire-PM2.5 decreased milk yield (MY) in a limited number of cows over one summer. The effects
of wildfire smoke on dairy cattle across multiple years and locations with varying wildfire smoke
conditions have not been assessed. The goal of this retrospective study was to expand current
knowledge on the impacts of wildfire smoke on lactating dairy cattle and calves by utilizing farm
records of milk production, morbidity, and mortality across multiple dairies in the Pacific Northwest.
Animal health and production data were collected for 3 dairies from their respective DairyComp305
systems. We used individual cow milk yield data from DHIA milk test days during wildfire months
(Jun 1 – Oct 31) for years when there was at least one wildfire smoke event (defined by 24-hr average
fine particulate matter (PM2.5) > 35µg/m3 [the air quality threshold used for wildfire smoke exposure).
Sample sizes and years included in the analyses for each farm were n = 10,709 for the years 2018 and
2020 on Dairy 1, n = 33 for 2018 on Dairy 2, and n = 501 for the years 2006, 2012, 2015, 2017, 2018,
2020, 2021, 2022, 2023 on Dairy 3. Dairy 3 also recorded daily milk yield (n =195) during the
wildfire smoke season for the years 2020-2023. For most test days, proximal milk components
[(MC), (total fat and total protein], and somatic cell count (SCC) were available for Dairies 2 and 3.
All-cause mortality, mastitis, and pneumonia cases were collected from all 3 dairies (1 & 2: 2018-
2020, 3: 2020-2023). Daily, 24-h average PM2.5 concentration and temperature humidity index (THI)
were calculated from state and national databases. HYSPLIT modeling was used to track PM2.5 to
wildfires. All cow and calf mortality and pneumonia cases, in addition to mastitis cases (cows only),
were plotted over time and visually compared against daily PM2.5 and THI. MY and MC data were
analyzed in SAS; MY and proximal MC for each farm were run with separate general linear mixed
models (PROC MIXED) for each level of parity (primiparous and multiparous) and lactation stage
(early-, mid- and late-lactation). All models included fixed effects of PM2.5, THI, and PM2.5 × THI
interactions (7-d averages and 7-d maximums), with cow as a random effect. Parity models included
days in milk (DIM) and lactation stage models included parity as additional fixed effects. Generalized
linear mixed models (PROC GLIMMIX) with negative binomial distribution and a log link were used
to analyze SCC. Daily MY from Dairy 3 were analyzed using separate general linear mixed models
for each parity level and lactation stage, with daily average PM2.5 and THI and their interaction as
fixed effects, cow as a random effect, and lags up to 7 days for PM2.5 and THI. On Dairy 1, combined
elevated PM2.5 and THI (both 7-d avg and 7-d max) decreased MY for primiparous and multiparous
cows in mid- and late-lactation (P <0.0001); but positively impacted MY for early lactating cows (P
<0.0001). On Dairy 2, elevated 7-d max PM2.5 and THI interactions increased MY for multiparous
cows (P = 0.004). Milk test day MY on Dairy 3 was positively impacted by an increased PM2.5 and
THI interaction for primiparous, early- and late-lactation cows (both 7-d avg and 7-d max) (P
<0.006), as well as multiparous cows (7-d average only) (P =0.007). Elevated 7-d max PM2.5 alone
decreased MY for multiparous animals on Dairy 3 (P =0.003). Daily MY on Dairy 3 decreased across
all 7-d lags for multiparous, early-, mid-, and late-lactation cows with combined elevated PM2.5 and
THI (P <0.0001). The MC and SCC response to wildfire smoke events varied by both farm and milk
component. While sample size was not sufficient for statistical analysis, cow and calf mortality and
pneumonia rates do not appear to be impacted on any farm, except for a slight increase in cow
pneumonia on Dairy 3, and an increase in calf pneumonia and mortality on Dairy 1 following a large
smoke event. While limited because of availability of farm records, this study increased the sample
size and diversity of cattle parity and lactation stages in current wildfire and dairy cattle research and
indicate that wildfire smoke has adverse effects on dairy cattle productivity and possibly health.