Hygroscopic weight gain of pollen grains from Juniperus species

Abstract

Juniperus pollen is highly allergenic and is produced in large quantities across Texas, Oklahoma, and New Mexico. The pollen negatively affects human populations adjacent to the trees, and since it can be transported hundreds of kilometers by the wind, it also affects people who are far from the source. Predicting and tracking long-distance transport of pollen is difficult and complex. One parameter that has been understudied is the hygroscopic weight gain of pollen. It is believed that juniper pollen gains weight as humidity increases which could affect settling rate of pollen and thus affect pollen transport. This study was undertaken to examine how changes in relative humidity affect pollen weight, diameter, and settling rate. Juniperus ashei, Juniperus monosperma, and Juniperus pinchotii pollen were applied to greased microscope slides and placed in incubation chambers under a range of temperature and humidity levels. Pollen on slides were weighed using an analytical balance at 2- and 6-h intervals. The size of the pollen was also measured in order to calculate settling rate using Stokes' Law. All pollen types gained weight as humidity increased. The greatest settling rate increase was exhibited by J. pinchotii which increased by 24 %.

Full text

ORIGINAL PAPER
Hygroscopic weight gain of pollen grains from Juniperus species
Landon D. Bunderson & Estelle Levetin
Received: 12 June 2013 /Revised: 26 June 2014 / Accepted: 27 June 2014
#
ISB 2014
Abstract Juniperus pollen is highly allergenic and is pro-
duced in large quantities across Te xas, Oklahoma, and New
Mexico. The pollen negatively affects human populations ad-
jacent to the trees, and since it can be transported hundreds of
kilometers by the wind, it also affects people who are far fro m
the source. Predicting and tracking long-distance transport of
pollen is difficult and complex. One parameter that has been
understudied is the hygroscopic weight gain of po llen. It is
believed that juniper pollen gains weight as humidity increases
which could affect settling rate of pollen and thus affect pollen
transport. This study was undertaken to examine how changes
in relative humidity affect pollen weight, diameter , and settling
rate. Juniperus ashei, Juniperus monosperma,andJuniperus
pinchotii pollen were applied to greased microscope slides and
placed in incubation chambe rs under a range of temperature
and humidity levels. Pollen on slides were weighed using an
analytical balance at 2- and 6-h intervals. The size of the pollen
was also measured in order to calculate settling rate using
Stokes’ Law. All pollen types gained weight as humidity in-
creased. The greatest settling rate increase was exhibite d by
J. pinchotii which increased by 24 %.
Keywords Hygroscopic
.
Pollen
.
Juniper
.
Settling rate
.
Terminal velocity
.
Mountain cedar
Introduction
The Cupressaceae is a significant source of airborne allergens, and
the genus Juniperus is a major component of many ecosystems
across the northern hemisphere (Mao et al. 2010; Pettyjohn and
Levetin 1997). New Mexico, Texas, and Oklahoma are home to
many species of juniper . Three species that represent a significant
allergy contribution are Juniperus ashei, Juniperus monosperma,
and Juniperus pinchotii. J. ashei pollen is considered the most
allergenic species of Cupressaceae in North America (Rogers and
Levetin 1998). This species is distributed throughout central
Texas, Northern Mexico, the Arbuckle Mountains of south central
Oklahoma, and the Ozark Mountains of northern Arkansas and
southwestern Missouri and pollinates from December to February
(Pettyjohn and Levetin 1997;Adams2008). J. monosperma is
distributed throughout south central Colorado, much of New
Mexico, Arizona, the pan handle of Oklahoma, and the pan
handle of T exas as well as south western T exas (Adams 2008)
J. monosperma is known to show cross-reactivity with J. ashei
(Schweitz et al. 2000); this species pollinates from late winter to
early spring (February–April) (Adams 2008). J. pinchotii is also
cross-reactive with J. ashei and endemic to the south central
United States, in central and western Texas, and adjacent
Southwest Oklahoma, New Mexico, and Coahulia Mexico
(Schweitz et al. 2000; Eckenwelder 2009). It is an important
aeroallergen which pollinates from late September to late
November (Pettyjohn and Levetin 1997; Weber and Nelson
1985; Wodehouse 1971).
The pollen from these species is produced in quantities of
up to 523 billion pollen grains per tree (Bunderson et al.
2012). In addition, the pollen grains are small and lightweight
and can be transported over great distances (Levetin and Buck
1986; Rogers and Levetin 1998;Levetin1998; Levetin and de
Water 2003). The distance that pollen travels after release
depends on the pollen type and meteorological conditions.
Although the majority of pollen released is deposited near
L. D. Bunderson (*)
Department of Agronomy, Iowa State University, Ames, IA 500114,
USA
e-mail: bundersonlandon@hotmail.com
E. Levetin
Department of Biology, The University of Tulsa, Tulsa, OK 74104,
USA
Int J Biometeorol
DOI 10.1007/s00484-014-0866-9

the source, a small percentage of pollen “escapes.” The per-
centage of pollen or spores that is transported high enough in
the atmosphere such that it is beyond the risk of immediate dry
deposition on the ground is known as the “escape fraction”
(Gregory 1978). Particles remain suspended as long as the
upward movement of air is faster than the terminal velocity of
the particle (Gregory 1973). The terminal velocity of
suspended particles is affected by the environmental condi-
tions of the air masses they encounter. For example, it is well
known that air temperature, humidity, and pressure can affect
the density and viscosity of air which can, in turn, affect
settling rate (Tsilingiris 2008;Gregory1978).
Relative humidity fluctuations can also affect pollen termi-
nal velocity due to hygroscopic weight gain. Change in weight
of pollen across a range of relative humidity levels has been
estimated for various species. For example, four Poa species
gained weight when exposed to a range of humidity levels
(Diehl et al. 2001). Often, in response to fluid loss, pollen
grains will accommodate change in cellular volume by folding
inward on apertures during desiccation and unfolding during
hydration (Wodehouse 1935). This folding and unfolding
mechanism is referred to as harmomegathy (Wodehouse
1935). Although juniper pollen grains do not have a furrow
and only have a very small pore, samples of the juniper pollen
grains collected in this study were observed at collection and
were in a non-spherical, desiccated state.
Pollen size, density, and settling rate have been estimated and
calculated using a variety of methods. For example, Gilissen
(1977) used a controlled humidity environment in an enclosed
chamber to expose pollen to a desired temperature and humidity
level and after which, the pollen was measured to determine the
effects of the exposure. Durham (1946) estimated the settling rate
of pollen by timing its travel through a 1-m tube. Settling rate is
also commonly estimated using Stokes Law (Gregory 1973).
Corn pollen density was calculated by measuring settling rate in
uniform fluids with known specific gravity and viscosity (van
Hout and Katz 2004). Once settling rate is established, pollen
density can be estimated using Stokes’ flow equation and this
method requires no prior knowledge of particle size (van Hout
and Katz 2004). Other methods require the measurement of
pollen size in order to estimate pollen density. One method
employs a gas pycnometer to determine particle density . A gas
pycnometer can be used to measure the volume of a sample of
many particles (e.g., pollen grains) by compressing a gas into the
container holding the particles. The volume occupied by the
particles is calculated based on the known properties of the gas
and the expected amount of gas in the chamber minus the actual
amount of gas in the chamber . This method has been used for
corn pollen and ragweed pollen but the method is limited be-
cause it only measures the solid portion of the pollen grain
(Harrington and Metzger 1963; Aylor 2002). Air that may have
seeped into the pollen grain, altering the size and shape of the
pollen, is not properly accounted for (van Hout and Katz 2004).
Another method for estimating particle density requires an elec-
trodynamic balance apparatus (EDB) which uses a net charge
and a synthetic air environment to suspend particles. An EDB
was used to measure the changing mass of Sailx, Betula,and
Nar cissus pollen (Pope 2010). The average increase in mass due
to water uptake at 75 % humidity was 16 % (Pope 2010).
There is no information on how changing relative humidity
levels affect the weight of Cupressacea e pollen. However, it is
known that the thick cellulose-rich intine of Cupressaceae pollen
is highly absorbent and will, in solution, absorb moisture and
swell until the thin inflexible exine is shed (Takaso and Owens
2008). This swelling and shedding actio n is an important step in
the pollination process when it occurs while suspended in the
pollination drop. The pollination drop is a solution secreted by the
ovule that acts as a kind of “liquid stigma,” collecting pollen from
the air (Dörken and Jagel 2014). Once the pollen has landed on a
pollination drop, fluid is thought to flow through a very small pore
in the exine which is plugged with a temporary structure, often
called an operculum (Duhoux 1982). As the intine swells, it sheds
the exine in a matter of minutes and depending on the solution,
the intine itself breaks h ours or days later releasing the protoplast
(Chichiriccò and Pacini 2008). On a compatible ovule when the
non-elastic exine is shed, the remaining portion of the pollen is
flexible and travels more easily through the micropyle and mi-
cropylar canal (T ak aso and Owens 2008). The absorption of water
by the intine could also affect the flight of airborne pollen. If water
vapor were absorbed through the exine it would affect the air-
borne transport of the pollen by changing the weight, size, and
shape of the pollen grains. Though the amount of weight gain is
unknown, Cupressus arizonica pollen has been shown to shed its
exine at 100 % relative humidity between 6 and 24 h (Chichiricco
et al. 2009).
Most pollen is released from anthers or microsporangiate
cones during the daytime, and pollen concentrations are often
negatively correlated with relative humidity (Weber 2003).
Glassheim et al. (1995) suggest that the reason for this is that
moist air inhibits pollen dispersal due to hygroscopic weight
gain and interference with floret drying and anther separation.
J. ashei pollen requires relative humidity below 50 % and
temperature above 5 °C with dry conditions persisting for 24 h
(Levetin and de Water 2003). Despite the required conditions
for release, juniper pollen can come into contact with a wide
variety of temperatures and humidity levels because it can be
suspended in the air for long periods (Van de Water and
Levetin 2001). Knowing how much hygroscopic weight gain
juniper pollen experiences across the range of possible hu-
midity levels could elucidate the degree of the humidity effect
on pollen settling rates and inform researchers interested in
conducting dispersion modeling studies. This experiment was
designed to test the magnitude of the effect of changing
relative humidity on pollen weight, size, and settling rate. It
is expected that as relative humidity increases, weight and size
will also increase.
Int J Biometeorol

Materials and methods
Pollen
J. ashei, J. monosperma,andJ. pinchotii pollen were collect-
ed from native populations and refrigerated at 4 °C until used.
The number of pollen grains per milligram was determined
using the hemocytometer method. Two milligrams of J. ashei
pollenwassuspendedin1.5mlofFAA,and2.5mg
J. monsperma and J. pinchotii pollen were suspended in
1 ml of FAA. Approximately 10 μl of the suspension was
placed on each side of the hemocytometer. The number of
pollen grains was estimated using standard hemocytometer
dilution conversions, and the counts were repeated six times
(Pettyjohn and Levetin 1997).
Humidity experiments
Pollen were exposed to different levels of relative humidity
inside desiccation chambers and evaluated after exposure
similar to that of Gilissen (1977)andChichiriccoetal.
(2009). Ten microscope slides (pollen slides) were coated with
Lubriseal stopcock grease (Thomas Scientific) and dusted
with pollen using a 1-ml syringe. Ten grease-coated but non-
pollen-dusted slides (greased) and ten non-greased, non-
dusted slides (ung reased) w ere used as controls. The
ungreased slides were used to visually track condensation on
the slides in order to eliminate condensation as a factor.
Treatments with apparent condensation were eliminated from
analysis and rerun.
Two 30×30×30 cm polyurethane desiccation boxes were
placed inside a climate-controlled growth chamber (Percival
Industries, Atlanta, GA) or a walk-in cold room (4 °C treat-
ment) and used to expose the pollen-coated slides to various
temperatures and humidity levels. In order to achieve a wide
range of humidity levels, two humidity control methods were
employed: dry silica gel and saturated salt solutions (Connor
and Towill 1993). For 20 and 40 % humidity levels (Art
Preservation Services, Long Island City, NY) and 50 and
60 % humidity levels (Heartfelt Industries), dry silica gel
was used. In order to achieve 76 % humidity, 300 ml of a
saturated NaCl solution (>0.37 g/ml) was used. For 85 %
humidity, 300 ml of saturated KCl (>0.40 g/ml) was used.
Ninety-seven percent humidity was achieved with 300 ml of a
saturated K
2
SO
4
(>0.12 g/ml) solution (Rockland 1960). Two
hygrometers were placed at different elevations inside both
desiccation boxes in order to monitor humidity.
Pollen-dusted slides were standardized by incubation at
20 °C and 60 % humidity for at least 12 h and then placed
in the second desiccation box for the desired experimental
humidity and temperature and measured at 2- and 6-h inter-
vals. The 2-h time interval was chosen because it was the
minimum time needed for the chambers to equilibrate for the
low humidity levels and 6 h was chosen because it was the
minimum time required for C. arizonica exineshedat100%
relative humidity (Chichiricco et al. 2009). New pollen slides
were created for each temperature and humidity level. All
slides were weighed using an analytical balance. Humidity
of the saturated salt solutions varied slightly with temperature.
Humidity treatments will be referred to in terms of the relative
humidity (RH) expected at 20 °C to eliminate confusion. A
detailed breakdown of actual humidity at a given temperature
and salt solution combination is provided (Table 1). J. ashei
and J. monosperma pollen were exposed to 20, 40, 50, 76, 85,
and 97 % humidity at 20 °C. J. ashei was also exposed to 20,
40, 76, and 97 % at 15 °C. J. pinchotii was exposed to 20, 40,
76, and 97 % at 20 °C. For J. ashei,anadditionaltemperature
of 4 °C was achieved in a walk-in cold room using humidity
levels of 20, 40, 50, 76, 85, and 97 % (Table 1). Limited pollen
inventory prevented all juniper species from being exposed to
all levels of humidity and temperature and from further repli-
cation. A pollen slide was analyzed after each 6-h humidity
exposure treatment in order to determine which, if any, hu-
midity levels caused the exine to be shed. This was achieved
by adding a drop of immersion oil and a cover slip to the
pollen slide (Pacini et al. 1999). At least 100 pollen grains
were counted in a traverse of the microscope slide.
Settling rate
Stokes Law was used to estimate the effect of the change in
humidity on settling rate (terminal velocity) of J. ashei pollen.
Stokes Law estimates the velocity of sedimentation (v
s
)fora
spherical particle: v
s
¼
2
9
P
p
−P
f
ðÞ
μ
gr
2
(Gregory 1973). The
density and viscosity of air varies with elevation, water con-
tent, and temperature. In order to complete the calculations,
the following variables were used: g (981 cm/s), P
p
(pollen
density (g/cm
3
)), P
f
(density of air (g/cm
3
)), and μ (viscosity
of air (g/cm/s )) as c alculated using the equations from
Tsilingiris (2008) and for simplicity, barometric pressure was
assumed to be 1 atm.
Table 1 Temperature and humidity levels for J. ashei, J. monosperma,
and J. pinchotii
Treatment J. ashei J.monosperma J.pinchotii
Rel. humidity 20 °C 15 °C 4 °C 20 °C 20 °C
20 % 20 % 20 % 20 % 20 % 20 %
40 % 40 % 40 % 40 % 40 % 40 %
50 % 50 % ND 50 % 50 % ND
76 % 76 % 76 % 75 % 76 % 76 %
85 % 85 % ND 88 % 85 % ND
97 % 97 % 99 % 98.5 % 97 % 97 %
ND no data available
Int J Biometeorol

Pollen diameter was determined using the following pro-
cedure: Three pollen-dusted slides were placed in 20 °C and
20 % RH conditions for 6 h, and 3 dusted pollen slides were
placed in 20 °C and 97 % RH conditions for 6 h. Both
humidity levels were repeated at 4 °C for J. ashei only.
After the allotted time, a drop of immersion oil was
added to the pollen on each slide (Pacini et al. 1999). A
cover slip was added and the slide was imme diately
photographed at ×400. A t least 100 pollen grains were
measured for each J. ashei treatment and 30 pollen
grains for each J. monosperma and J. pinchotii treat-
ment. Images of pollen were analyzed to determine
pollen cross-sectional area using ImageJ, and diameter
was calculated using d ¼
ffiffiffiffiffiffiffiffiffiffiffiffiffi
4A
p
=π
p
where A
p
is the
projected are a of the pollen g rain (van Hou t et al.
2008). Pollen volume was also calculated and combined
with pollen weight values to determine pollen density.
Statistical analysis
A two-way repeated measures ANOVA analysis was
performed for J. ashei, J. pinchotii,and
J. monosperma using time as the repeated factor and
slide treatment (pollen slide/greased control/ungreased
control) and humidity as additional factors. In the case
of J. ashei, temperature was an additional factor. Since
not all levels of humidity w ere used for t he three levels
of temperature, temperature was analyzed separately in a
two-way A NOVA using the levels of temperature and
slide treatment. Humidity and pollen weight gain were
also tested independently using simple regression. In
addition , a t test was used to compare pollen diameter
at 97 % RH and 20 % RH for J. monosperma and
J. pinchotii and a two-way ANOVA was used to com-
pare the J. ashei treatments (SAS JMP 2010).
Results
The average weight per pollen grain as measured using a
hemocytometer after incuba tion at 60 % RH was 2.5×
10
−6
mg, 4.0×10
−6
mg, and 4.8×10
−6
mg for J. ashei, J
pinchotii, and J. monosperma, respectively. Pollen weight
was positively correlated with humidity for all species and at
all temperatures (Figs. 1, 2,and3). In addition, weight change
was usually negligible between 2- and 6-h intervals. Exine
shed was also negligible. No less than 98 % of pollen grains
had their exines intact before humidity treatment for each
juniper species at any given humidity level, and the same
percentage existed after the 6-h treatment. If the high humidity
treatments caused exine shed at the 6-h juncture, our sample
size was not large enough to detect this.
Juniperus ashei
At 20 °C, J. ashei pollen weight was significantly affected by
humidity level (F
1,174
=55.9, p<0.0001), slide treatment
(F
2,174
=8.19, p<0.0005), and the humidity×slide treatment
interaction (F
2,174
=76.1, p<0.0001). As expected, the slide
treatment and the slide treatment×humidity interaction were
significant factors affecting weight gain. The control slides’
treatments lacked pollen; thus, there was no weight gain due to
water absorption. Incubation time was not a significant factor
affecting pollen weight (F
1,174
=0.09, p=0.77) nor were any of
the other interaction factors (time×humidity F
1,174
=1.32, p=
0.25; time×slide treatment: F
2,174
=0.77, p=0.46; time×hu-
midity×slide treatment: F
2,174
=0.28, p=0.76).
J. ashei pollen weight was significantly affected by humid-
ity level at 15 °C (F
1,113
=64.17, p<0.0001). Additionally,
weight was affected by slide treatment and the interaction
humidity×slide treatment (respectively, F
2,113
=5.81,
p<0.005 and F
2,113
=114.62, p<0.0001). Time and the re-
maining interaction effects were not significant factors (time:
F
1,113
=0.12, p=0.73; time×humidity F
1,113
=0.66, p=0.42;
time×slide treatment: F
2,113
=0.41, p=0.67; time×humidi-
ty×slide treatment: F
2,113
=0.66, p=0.52). The pollen-dusted
Fig. 1 Average weight per pollen grain of J. ashei at 2-h reading plotted
against VPD with logarithmic fit line (p<0.0001) for temperature (20, 15,
and 4 °C)×humidity (20, 40, 50, 76, 85, 97 %) combinations
Fig. 2 Average weight per pollen grain of J. pinchotii at 20 °C and the
2-h reading plotted against VPD with logarithmic fit line (p<0.0001)
Int J Biometeorol

slide weight fluctuated with increased humidity which is why
the slide treatment and the interaction humidity×slide treat-
ment were significant.
For the J. ashei 4 °C experiment, humidity and the
interaction humidity×slide treatment were not significant
factors affecting weight cha nge (respectively, F
1,173
=
0.72, p=0.40 and F
2,173
=0.58, p=0.56) but the interac-
tions tim e×humidity, time×slide t reatm ent, and t ime×
humidity×slide treatment were significant (respectively,
F
1,173
=10.59, p <0.005, F
2,173
=9.18, p <0.0005, and
F
2,173
=4.02, p <0.05). The weight of pollen-dusted
slides in the 4 °C 97 % hum idity treatment continued
to increase with time causing the abovementioned inter-
actions that involved time. Slide treatment and time
were not significant factors (respectively, F
2,173
=1.01,
p=0.37 and F
1,173
=2.02, p=0.16).
ThechangeinweightforJ. ashei pollen across the range of
humidity treatments was similar for all three temperature
levels. Estimated average weights after 2 h for the 97 %
humidity treatment at 4, 15, and 20 °C were, respectively,
2.77×10
−6
mg, 2.76×10
−6
mg, and 2.79×10
−6
(Fig. 1).
However, the 6-h mean weight for the 4 °C, 97 % humidity
treatment was 2.91×10
−6
mg while the 97 % humidity treat-
ments for 15 and 20 °C only deviated slightly from 2-h
readings. In spite of the deviation at the 6-h reading for the
4 °C experiment, the two-way repeated measures ANOVA for
the three temperature treatments for J. ashei showed that
temperature did not have a significant effect on pollen weight
(F
1,476
=0.23, p=0.63).
In order to examine weight change across the range of
humidity levels and temperatures, vapor pressure deficit
(VPD) values were calculated (Table 2) and a regression
analysis was used to test the relationship be tween pollen
weight and VPD (Fonseca and Westgate 2005). A logarithmic
fit yielded an R
2
value of 0.66 (p<0.0001) for the 2-h weights
(Fig. 1). Since time was not a significant factor affecting
J. ashei pollen weight, only 2-h readings are plotted (logarith-
mic fit for 6-h weights R
2
=0.62, p<0.0001).
Juniperus pinchotii
Humidit y, slid e treat men t , and the interac ti on hum idit y×
slide treatment were significant factors affecting weight
change for J.pinchotii (respectively, F
1,114
= 71.5,
p < 0.0001, F
2,114
=3.8, p <0.05, and F
2,114
= 43.5,
p<0.0001). Time and the interaction time×slide treat-
ment also significantly affected weight (F
1,114
=4.7,
p <0.05 and F
2,114
=4.1, p<0.05), but the remaining
interaction effects w ere not significant factors (time×
humidity: F
1,114
=0.08, p=0.78; time×slide treatment×
humidity: F
2,114
=0.46, p=0.63). As mentioned previous-
ly, the humidity×slide treatment interaction is a n ex-
pected result because the control slides should not be
affected by humidity and pollen gains weight due to
water a bsorption. The time and t ime interaction effect
was in part due to weight fluctuation at the different
time intervals for the pollen-dusted slides at the 7 6 and
97 % humidity treatments. The 2-h mean weight was
slightly heavier than the 6-h weight for the 97 % hu-
midity treatment and vice versa for the 76 % humidity
treatment. A simple regression using weight and VPD
shows a positive, significant correlation between humid-
ity and pollen weight for the 2-h time interval (Fig. 2)
as well as the 6-h time i nterval (not shown p<0.0001,
R
2
=0.73).
Juniperus monosperma
For J. monosperma, the slide treatment×humidity interaction
was the only significant factor affecting weight change
(F
2,174
=4.0, p<0.05). Humidity, slide treatment, time, and
the interactions, time×humidity, time×slide treatment, and
time×humidity×slide treatment were not significant factors
affecting weight change (respectively, F
1,174
=0.002, p=0.96,
F
2,174
=1.78, p=0.17, F
1,174
=0.21, p=0.65, F
2,174
=0.92, p=
0.40, F
1,174
=0.88, p=0.35and F
2,174
=0.75, p=0.47). The ma-
jor effect was that the weight of the pollen-dusted slides
increased substantially more than control slides across the
range of relative humidity levels. Simple regression using
pollen weight and VPD shows a positive, significant correla-
tion between humidity and pollen weight for the 2-h time
Fig. 3 Average weight per pollen grain of J. monsperma at 20 °C and the
2-h reading plotted against VPD with logarithmic fit line (p<0.0001)
Table 2 Vapor pressure
deficit values (hPa) for
temperature and relative
humidity combinations
Temp/RH 20°C 15°C 4°C
20 % 18.70 13.64 6.50
40 % 14.03 10.23 4.88
50 % 11.69 8.53 4.07
76 % 5.61 4.09 2.03
85 % 3.51 2.56 0.98
97 % 0.70 0.17 0.12
Int J Biometeorol

interval (Fig. 3)aswellasthe6-htimeinterval(notshown
p<0.0001, R
2
=0.77).
Diameter/settling rate
Mean diameter of J. ashei pollen exposed to 20 % RH and
20 °C was 18.51 μm, and th e mean diameter for pollen
exposed to 97 % RH and 4 °C was 19.36 μm(Table3). For
J. monosperma, mean diameter was 23.9 for 20 % RH and
24.1 μm for 97 % RH. J. pinchotii diameters were 21.2 for
20 % RH and 21.6 μm for 97 % RH. Changes in diameter
sizes were not significant for J. monosperma and J. pinchotii
(t=0.013 p=0.910 and t=0.005 p=0.944, respectively). The
two-way ANOVA was significant for J. ashei pollen diameter
(F
3,424
=10.13, p<0.0001). Humidity and temperature were
significant factors affecting J. ashei diameter size (respective-
ly, F
1,212
=12.07, p<0.0006; F
1,212
=17.86, p<0.001). The
interaction humidity×temperature was not significant
(F
1,212
=0.46, p=0.498).
Pollen weight change was not proportional to pollen diam-
eter change from 20 to 97 % RH for any of the three pollen
types which meant that density of the pollen grains increased
with the increase in h umidity. This had an effect on the
calculated settling rate (Table 3). Although the change in
weight was not proportional to the change in size for any of
the species of pollen, the greatest change in settling velocity
was correlated with the pollen type with greatest percent
change in weight (Table 3).
The calculated settling rates showed varying levels of
change across species, temperatures, and humidity. A change
in humidity had a greater effect on settling rate than a change
in temperature (Table 3). The greatest changes in settling rates
for J. ashei, J. monosperma,andJ. pinchotii,were16,18,and
24 %, respectively (Table 3).
Discussion
The estimated average weight per pollen grain for J. ashei at
60 % RH was 2.5×10
−6
mg in this study, but Pettyjohn and
Levetin (1997) calculated the average weight of a J. ashei
pollen grain to be 4.6×10
−6
mg. The mean diameter of the
J. ashei pollen in this study was between 18.51 and 19.36 μm
depending on temperature and humidity. Wodehouse (1935)
reported that dry pollen sizes of J. ashei ranged from 18.2 to
21.6 μm in diameter and moist grains ranged from 20.5 to
22.8 μm in diameter. It is possible that the weights from this
study differ from that of Pettyjohn and Levetin (1997) due to a
difference in mean pollen size. The J. ashei pollen used in this
study was collected from a single location in central Texas
near the town of Lampasas. Pollen observed in some other
locations was larger than the Lampasas pollen (data not
shown). It is not clear whether sizes vary widely from cone
to cone, tree to tree, or location to location.
Response to relative humidity at the chosen time intervals
was relatively uniform. In most cases, virtually all of the size
and weight change happened in the first 2 h. These changes
may have occurred much faster than 2 h. Unfortunately, since
it can take several minutes for a humidity chamber to equili-
brate, shorter time intervals were not possible, especially for
lower relative humidity treatments. This means that actual
response time cannot be measured. Nevertheless, the experi-
ment shows the range in possible changes. Pollen weight
change response time due to changing relative humidity re-
mains to be tested on anemophilous pollen.
One result that may represent a deviation from the response
of a related species is the lack of exine shed. Although
C. arizonica pollen grains shed their exines at 100 % relative
humidity between 6 and 24 h of exposure (Chichiricco et al.
2009), exine shed in this study was negligible. This study used
97 % relative humidity and 6 h was the maximum time for
Table 3 Density and viscosity of air and settling rate of J. ashei, J. monosperma, and J. pinchotii pollen at two temperatures and two levels of relative
humidity
Temp °C/RH% VPD (hPa) Density of air g/cm
3
Viscosity of air g/cm s Diameter μm (SD) Weight ng (SD) Density g/cm
3
Settling rate cm/s
J. ashei
20/97 0.70 1.198×10
−3
1.85×10
−4
18.97 (2.26) 2.87 (0.18) 0.79 0.83
20/20 18.70 1.204×10
−3
1.83×10
−4
18.51 (2.50) 2.37 (0.07) 0.72 0.73
4/97 0.12 1.271×10
−3
1.76×10
−4
19.36 (2.03) 2.91 (0.19) 0.77 0.88
4/20 6.50 1.274×10
−3
1.75×10
−4
19.06 (2.07) 2.45 (0.06) 0.68 0.76
J. monosperma
20/97 0.70 1.198×10
−3
1.85×10
−4
24.08 (2.63) 5.55 (0.10) 0.76 1.29
20/20 18.70 1.204×10
−3
1.83×10
−4
23.94 (2.77) 4.58 (0.17) 0.64 1.09
J. pinchotii
20/97 0.70 1.198×10
−3
1.85×10
−4
21.59 (2.98) 4.63 (0.29) 0.88 1.24
20/20 18.70 1.204×10
−3
1.83×10
−4
21.61 (3.22) 3.78 (0.12) 0.72 1.00
Int J Biometeorol

exposure. It is not clear whether the reason for the discrepancy
in exine shed is due to differences in species or tim e of
exposure. Since Lubriseal is a non-water-based substance
and the pollen was resting on the surface of the grease, it is
not anticipated that the Lubriseal had any effect on the hygro-
scopic experiment or exine shedding. It is possible that 100 %
relative humidity treatments by Chichiricco resulted in water
droplets caused by condensation—which would mean that the
pollen grains were suspended in water—resulting in the dis-
solution of the operculum. The J. ashei pollen grains used in
this study were observed suspended in solutions of different
water activity levels (data not shown), and exines were often
rapidly shed in these conditions.
Percent weight change for the juniper species was in a
similar range to that of Salix, Betula,andNarcissus as reported
by Pope (2010). The greatest changes in weight for J. ashei,
J. monosperma, and J. pinchotii, were 25, 21, and 24 %,
respectively, while Pope reported average increase in mass
due to water uptake at 75 % humidity to be 16 % (Pope 2010).
Unfortunately, since Pope’s reported range was from 0 to
75 %, a direct comparison cannot be made (Pope 2010).
Durham (1946) estimated the settling rate of pollen by
timing its travel through a 1-m tube. He found the rate of fall
for J. ashei to be 1.16 cm/s which is faster than the fastest
J. ashei-calculated speed from this study. However, Durham
(1946) reported that the mean diameter of the J. ashei was
22.8 μm. This is closer to J. pinchotii and J. monosperma
diameters, and the 1.16 cm/s rate does fit within their range of
settling rates (Table 3). Also, comparisons of observed settling
rates to calculated settling rates have often been found to be
quite different. Durham (1946) suggested that the design,
specifically the diameter, of the chamber used to observe the
fall of pollen can affect the speed of fall. Also, static charge on
pollen grains can create large clumps of pollen which fall
faster than individual grains creating a downdraft and thus
increasing the rate of fall in individual pollen grains but
Durham (1946) tried to remove static charge.
A common scenario in the J. ashe i pollination period
would be daytime temperatures around 20 °C with 20 % RH
and nighttime temperatures around 4 °C with very high rela-
tive humidity. This scenario does produce differences in set-
tling rates; however, the differences are minimal compared to
the expected fluctuations in vertical wind speed. The fastest
settling rate was that of J. pinchotii with a rate of 1.24 cm/s. At
that speed, and in the absence of vertical winds, it would still
take about an hour for pollen suspended 50 m above the earth
to reach the ground. Although the settling rate of pollen can
change dramatically with increase in humidity, the system that
transports pollen is complex and it is difficult to make broad
statements about how weight gain would affect pollen trans-
port. The intent of this study was to evaluate the magnitude of
the effect of changing relative humidity on pollen weight, size,
and settling rate. Raw data were intended to serve as inputs for
a NASA-funded juniper pollen forecast system. This study
was not intended to predict actual or relative pollen transport
distances without the aid of dispersion modeling.
Acknowledgments This work is funded through NASA’s Applied
Sciences, Public Health program proposal called “Decision support
through Earth Science Research Results” element of the NASA ROSES
2008 omnibus solicitation.
References
Adams RP (2008) Juniperus of Canada and the United States: taxonomy,
key distribution. Phytologia 90:3
Aylor DE (2002) Settling speed of corn (Zea mays) pollen. J Aerosol Sci
33(11):1601–1607
Bunderson LB, Van de Water P, Wells H, Levetin E (2012) Predicting and
quantifying pollen production in Juniperus ashei forests. Phytologia
94(3):417–438
Chichiriccò G, Pacini E (2008) Cupressus arizonica pollen wall zonation
and in vitro hydration. Plant Syst Evol 270(3–4):231–242
Chich iricco G, Spano L, Torraca G, Tartarin i A (2009) Hydration,
sporoderm breaking and germination of Cupressus arizonica pollen.
PlantBiol11:359–368
Connor KF, Towill LE (1993) Pollen-handling protocol and hydration/
dehydration characteristics of pollen for application to long-term
storage. Euphytica 68(1–2):77–84
Diehl K, Quick C, Matthias-Maser S, Mitra SK, Jaenicke R (2001) The
ice nucleating ability of pollen Part I: Laboratory studies in deposi-
tion and condensation freezing modes. Atmos Res 58:75–87
Dörken VM, Jagel A (2014) Orientation and withdrawal of pollination
drops in Cupressaceae sl (Coniferales). Flora-Morphology,
Distribution, Functional Ecology Plants 209(1):34–44
Duhoux E (1982) Mechanism of exine rupture in hydrated taxoid type of
pollen. Grana 21(1):1–7
Durham OC (1946) The volumetric incidence of atmospheric allergens
III: rate of fall of pollen grains in still air. J Allergy 17(2):70–78
Eckenwelder JE (2009) Conifers of the world. Timber Press, Portland
Fonseca AE, Westgate MA (2005) Relationship between desiccation and
viability of maize pollen. Field Crop Res 94(2):114–125
Gilissen LJW (1977) The influence of relative humidity on the swelling
of pollen grains in vitro. Planta 137:299–301
Glassheim JW, Ledoux RA, Vaughan TR, Damiano MA, Goodman DL,
Nelson DS, Weber RW (1995) Analysis of meteorologic variables
and seasonal aeroallergen pollen counts in Denver, Colorado. Ann
Allergy Asthma Immunol 75(2):149–156
Gregory PH (1973) The microbiology of the atmosphere 2nd edition.
Halstead Press, New York
Gregory PH (1978) Distribution of airborne pollen and spores and their
long distance transport. Pure Appl Geophys 116(2–3):309–315
Harrington JB, Metzger K (1963) Ragweed pollen density. Am J Bot
50(6):532–539
Levetin E (1998) A long-term study of winter and early spring tree pollen
in the Tulsa, Oklahoma atmosphere. Aerobiolgia 14:21–28
Levetin E, Buck P (1986) Evidence of mountain cedar pollen in Tulsa.
Ann Allergy 56:295–299
Levetin E, de Water V (2003) Pollen count forecasting. Immunol Allergy
Clin 23:423–442
Mao K, Hao G, Lie J, Adams RP, Milne RI (2010) Diversification and
biogeography of Juniperus (Cupressaceae): variable diversification
rates and multiple intercontinental dispersals. New Phytol 188:254–
272
Int J Biometeorol

Pacini E, Franchi CG, Ripaccioli E (1999) Ripe pollen structure and histo-
chemistry of some gymnosperms. Plant Systemat Evol 217:81–99
Pettyjohn ME, Levetin E (1997) A comparative biochemical study of
conifer pollen allergens. Aerobiologia 13:259–267
Pope FD (2010) Pollen grains are efficient cloud condensation nuclei.
Environ Res Lett 5 (6 pp)
Rockland LB (1960) Saturated salt solutions for static control of relative
humidity between 5° and 40 °C. Anal Chem 32(10):1375–1376
Rogers CA, Levetin E (1998) Evidence of long-distance transport of
mountain cedar pollen into Tulsa, Oklahoma. Int J Biometeorol
42:65–72
Schweitz LA, Goetz DW, Whisman BA, Reid MJ (2000) Cross-reactivity
among conifer pollens. Ann Allergy Asthma Immunol 84:87–93
Takaso T, Owens JN (2008) Significance of exine shedding in
Cupressaceae-type pollen. J Plant Res 121:83–85
Tsilingiris PT (2008) Thermophysical and transport properties of humid
air at temperature range between 0 and 100 °C. Energ Convers
Manage 49:1098–1110
Van de Water PK, Levetin E (2001) Contribution of upwind pollen
sources to the characterization of Juniperus ashei phenolog y.
Grana 40:133–141
van Hout R, Katz J (2004) A method for measuring the density of
irregularly shaped biological aerosols such as pollen. J Aerosol Sci
35:1369–1384
van Hout R, Chamecki M, Brush G, Katz J, Parlange MB (2008) The
influence of local meteorological conditions on the circadian rhythm
of corn (Zea mays L.) pollen emission. Agr Forest Meteorol 148:
1078–1092
Weber RW (2003) Meteorologic variables in aerobiology. Immunol
Allergy Clin 23:411–422
Weber RW, Nelson HS (1985) Pollen allergens and their inter-
relationships. Clin Rev Allergy 3:291–319
Wodehouse RP (1935) Pollen grains: their structure, identification and
significance in science and medicine. McGraw-Hill Book Company,
Inc, New York
Wodehouse RP (1971) Hayfever plants, 2nd edn. Hafner, New York
Int J Biometeorol