Annals of Botany 134: 43–57, 2024

https://doi.org/10.1093/aob/mcae029, available online at www.academic.oup.com/aob

ORIGINAL ARTICLE

Interactions of moisture and light drive lichen growth and the response to
climate change scenarios: experimental evidence for Lobaria pulmonaria
Martine Borge and Christopher J. Ellis*

Royal Botanic Garden Edinburgh, 20A Inverleith Row, Edinburgh, EH3 5LR, UK
*
For correspondence. E-mail c.ellis@rbge.org.uk

Received: 6 November 2023 Returned for revision: 12 February 2024 Editorial decision: 22 February 2024 Accepted: 1 March 2024

• Background and Aims There is growing interest in the functional ecology of poikilohydric non-vascular
photoautotrophs (NVPs), including ‘cryptogamic’ bryophytes and lichens. These organisms are structurally im-
portant in many ecosystems, contributing substantially to ecosystem function and services, while also being sen-
sitive to climate change. Previous research has quantified the climate change response of poikilohydric NVPs
using predictive bioclimatic models with standard climate variables including precipitation totals and temperature
averages. This study aimed for an improved functional understanding of their climate change response based on
their growth rate sensitivity to moisture and light.
• Methods We conducted a 24-month experiment to monitor lichen hydration and growth. We accounted for
two well-known features in the ecology of poikilohydric NVPs, and exemplified here for a structurally dominant
lichen epiphyte, Lobaria pulmonaria: (1) sensitivity to multiple sources of atmospheric moisture including rain,
condensed dew-formation and water vapour; and (2) growth determined by the amount of time hydrated in the
light, driving photosynthesis, referred to as the Iwet hypothesis.
• Key Results First, we found that even within an oceanic high-rainfall environment, lichen hydration was better
explained by vapour pressure deficit than precipitation totals. Second, growth at a monthly resolution was posi-
tively related to the amount of time spent hydrated in the light, and negatively related to the amount of time spent
hydrated in the dark.
• Conclusions Using multimodel averaging to project growth models for an ensemble of future climate change
scenarios, we demonstrated reduced net growth for L. pulmonaria by the late 21st century, explained by extended
climate dryness and lichen desiccation for periods when there is otherwise sufficient light to drive photosynthesis.
The results further emphasize a key issue of photoperiodism when constructing functionally relevant models to
understand the risk of climate change, especially for poikilohydric NVPs.

Key words: Climate change, epiphyte, hydration, lichen, non-vascular photoautotrophs, photoperiodism,
poikilohydric, thallus growth, vapour pressure deficit.

INTRODUCTION and communities are sensitive to spatial patterns of climate vari-

ability. This has been tested for lichens that are climatically syn-
There is growing interest in the functional ecology of algae, chronous on disjunct landmasses (Braidwood and Ellis, 2012),
bryophytes, lichens and cyanobacteria, which were traditionally explaining the distribution of both species (Giordani and Incerti,
described as ‘cryptogams’ (Smith, 1938a, b), but which also 2008; Root et al., 2014; McCune et al., 2022) and different func-
form an ecological grouping of ‘non-vascular photoautotrophs’ tional guilds (McCune et al., 2022; Phinney et al., 2022). As a
(NVPs) (Porada et al., 2023). A common feature is that these corollary, lichen diversity is expected to be impacted by current
organisms are poikilohydric, with the ability for direct uptake and future human-induced climate change (Ellis et al., 2007a,
of moisture from the atmosphere, but without homeostatic regu- b; Allen and Lendemer, 2016; Nascimbene et al., 2016; Rubio-
lation of their water status being susceptible to ambient wet- Salcedo et al., 2017; Smith et al., 2019).
ting–drying cycles (Green and Lange, 1995; Green et al., 2011). Climate change risk assessments targeted to lichens are – in
They nevertheless include species that are structurally important common with many other ecological groups – often based on
in many ecosystems, from deserts, wetlands and tundra to forest bioclimatic models (alternatively referred to as habitat/spe-
canopies, contributing significantly to primary production cies distribution/niche models, see Ellis, 2019a, and references
and nutrient cycling (Elbert et al., 2012; Porada et al., 2013, therein), in which a species’ recorded distribution is compared
2014), while also being sensitive to climate change and there- to climatic conditions at a baseline (temperature, precipitation,
fore having ecosystem consequences that are as yet unresolved etc.), then to future climate change scenarios to quantify the
(Porada et al., 2023). Taking the lichen symbiosis as an ex- spatial shift and/or loss of suitable climate space. The shifts in
ample, there is substantial evidence to show that lichen species
© The Author(s) 2024. Published by Oxford University Press on behalf of the Annals of Botany Company.
All rights reserved. For commercial re-use, please contact reprints@oup.com for reprints and translation
rights for reprints. All other permissions can be obtained through our RightsLink service via the Permissions
link on the article page on our site—for further information please contact journals.permissions@oup.com.
44 Borge and Ellis — Moisture, light and the climate change response of L. pulmonaria

a species’ suitable climate space can help to infer threat: (1) if predictive framework that links their hydration status with the
the species is thought unable to migrate to track suitable cli- ambient light regime, to explain physiological responses and
mate space, such as when habitat fragmentation weakens the growth. Our second aim was therefore to explain lichen growth
ecologically relevant dispersal rate (Schwartz, 1993; Travis, by integrating hydration with both seasonal and diurnal fluc-
2003); (2) if the population is thought unable to adapt to the tuations in the light regime for a given latitude, exploring the
changing in situ conditions, because of small population sizes consequences of photoperiodism under baseline and climate
and low genetic diversity (cf. Frankham, 1996; Hoffman and change projections.
Sgrò á2011); and/or (3) if individuals are thought unable to suf-
ficiently acclimate through phenotypic or physiological plas-
ticity (Matesanz et al., 2010; Nicotra et al., 2010). Exposure MATERIALS AND METHODS
to a shift and/or loss of suitable bioclimatic space can help in
assessing the extent to which migration, adaptation or acclima- Thallus calibration
tion is required for survival, while limits to these processes rep-
resent a species vulnerability (Ellis, 2013). Although a useful Twenty-four juvenile and non-reproductive thalli of L. pulmon-
framework, there is a well-established need to advance this cor- aria (cf. Scheidegger et al., 1998; Eaton and Ellis, 2014) were
relative bioclimatic approach towards an improved functional sampled from a natural donor population (native semi-natural
basis. It has been shown that integrating physiological response oak–hazelwood at Loch Barnluasgan, Scotland, 56.062°N,
and demographic consequences can achieve increased model 5.549°W: Fig. 1) and air-dried (21 °C, relative humidity ~40 %)
prediction accuracy at ecological scales (see examples provided to calculate their dry weight (DW). Thalli were split into two
by Keith et al., 2008; Fordham et al., 2013; Swab et al., 2015). batches of 12. Individual thalli were wetted to full hydration
Since they are poikilohydric (Green et al., 2010; Palmqvist et by repeat spraying with distilled water, and gentle shaking
al., 2010), the ambient moisture regime is one of the key param- to remove surface drops. Following protocols presented by
eters controlling lichen physiology; lichens will desiccate to be- Coxson, (1991) (see also Renhorn et al., 1997; Jonsson et al.,
come dormant when the environment is dry, becoming hydrated 2008), four thalli representing a subset of each batch were then
and physiologically active when the environment is wet. This clamped by two crocodile clips, positioned ~1 cm apart, which
forms the basis of the Iwet hypothesis (Palmqvist and Sundberg, were linked via a stereo cable to a HOBO U12 four-channel
datalogger (Onset Corp., MA, USA). A direct current of 2.5 V
2000; Palmqvist et al., 2010), which states that – with respect
was used to measure electrical conductance across the thalli at
to a light compensation point – growth depends on the propor-
1-min intervals and as the thalli dried under laboratory room
tion of time that a lichen is both hydrated and light-exposed,
temperature (21 °C, relative humidity ~40 %). Simultaneously,
enabling photosynthesis, though noting that night-time respir-
the remaining eight thalli from the same batch were weighed at
ation can also benefit growth (Bidussi et al., 2013; Alam et al.,
time zero, and at intervals of 5, 10, 20, 30, 40, 50, 60, 80, 100,
2015). The Iwet hypothesis poses two key challenges to standard
120, 180 and 300 min to estimate their real-time weight (W).
bioclimatic modelling for lichens, and poikilohydric NVPs
The experiment was cycled across the three subsets within each
more generally, which we address in this study.
batch (4 thalli × 3 subsets = 12 thalli per batch), and repeated
The first challenge is with respect to hydration. Precipitation
for the two batches, to create six longitudinal datasets in which
totals feature in the most widely used baseline and climate
change scenarios (Hijmans et al., 2005; Fick and Hijmans, an average for thallus conductivity was related to an average
2017; Karger et al., 2017) and have been a standard variable for thallus water content (WC, mgH2O per mg dry mass), cal-
culated as proportional to the dry weight:
in a majority of lichen bioclimatic models (see table 1 in Ellis,
2019a). However, lichen hydration is known to be controlled W − DW
by variables other than precipitation (Gauslaa, 2014), including WC =
DW
dew-formation when condensation occurs, and water vapour
(Phinney et al., 2018; Hovind et al., 2020). Our first aim was To create a standard predictive curve, electrical conductance
therefore to explore how the hydration of a key lichen epiphyte, was plotted against WC and fitted with an appropriate deter-
Lobaria pulmonaria, which can be a dominant species in tem- ministic function that normalized residuals while explaining a
perate forest canopies (James et al., 1977; Ellis et al., 2015), high degree of variance.
might be explained not only by precipitation but by alternative
sources of moisture, with a focus on vapour pressure deficit
Thallus hydration
(VPD). Integrating temperature and relative humidity, VPD is
relevant to both the availability of water vapour for lichen hy- Thallus hydration and climate variables were measured under
dration, and the potential for condensation, as well as to the experimental field conditions for the period June 2018 to July
evaporative drying effect of the atmosphere (Anderson, 1936; 2020, though allowing breaks required for equipment mainten-
Barry and Blanken, 2016). VPD provides useful information ance. Two juvenile and non-reproductive thalli of L. pulmon-
about how the hydration status of poikilohydric organisms aria were sampled from the donor population (see ‘Thallus
might be affected by moisture sources other than precipitation. calibration’, above) and air-dried and weighed. Following
The second challenge is with respect to the Iwet hypothesis. McCune et al., (1996) and Muir et al., (1997), the thalli were
The challenge of photoperiodism (Saikkonen et al., 2012; attached to a loop of 6.8-kg breaking strain monofilament line
Ettinger et al., 2021) recognizes that while climate variables are with a bird-ringing tag for identification, using a small patch
dynamic and are changing, the latitudinal gradient in seasonal of non-toxic aquarium-grade silicone at the thallus base. The
light availability is relatively static. A necessary advance in bio- thalli were suspended from growth frames (cf. Ellis et al., 2017;
climatic modelling for poikilohydric organisms is to develop a Ellis, 2020), each thallus being clamped by two crocodile clips,
 Borge and Ellis — Moisture, light and the climate change response of L. pulmonaria 45

Benmore Botanic Garden

Fig. 1. The location of the experimental site at Benmore Botanic Garden: (A) Scotland positioned within the UK, (B) an oceanic region of south-west Scotland
and (C) the location of the experimental site.

positioned ~1 cm apart, linked to a HOBO U12 four-channel per month (mean ± 1 s.e.), alongside the respective values for
datalogger (as above) mounted onto the frame. A direct current total monthly rainfall. Furthermore, it was our aim to quan-
of 2.5 V was used to measure electrical conductance across the tify the importance of different moisture sources in explaining
thalli at 2-h intervals. Thalli were replaced by new donors at L. pulmonaria hydration. We first used Spearman’s rank cor-
a maximum of 6-month intervals. Electrical conductance was relation to compare VPD with the respective daily rainfall
converted to thallus WC using the laboratory calibration curve amount. Second, we compared thallus WC (response) scored as
(as above), and an average for the two thalli formed an estimate being above (1) or below (0) a threshold value (WC ≥ 0.2; see
of L. pulmonaria hydration in real-time. ‘Lichen growth’, below), to the concurrent VPD and respective
The growth frames were located in the same landscape pos- daily rainfall amount (fixed effect predictors), using a gener-
ition (same aspect and slope) at Benmore Botanic Garden (Fig. alized linear model (GLM) with a binomial structure (Zuur et
1) though ~50 m distant from a meteorological station which al., 2009; Crawley, 2013). We simplified the full model using
measured rainfall (mm) per 24-h period, from 10 h each morning. backwards selection and calculated a percentage for the total
The growth frames also included two hygrochron iButton deviance-explained for an optimized model, and the deviance-
dataloggers (Analog Devices, Wilmington, NC, USA) which re- explained separately by each fixed effect. GLM was imple-
corded temperature (°C) and relative humidity (%) at 2-h inter- mented in R (R Development Core Team, 2020).
vals. Measurements of temperature and relative humidity were
averaged across the two dataloggers, and VPD was calculated: Lichen growth
Å ã  
100 − rH 7.5T
Twenty additional juvenile and non-reproductive thalli of
VPD (Pa) = × 610.7 × 10( 237.3+T )
100 L. pulmonaria were sampled from the donor population and
Since we were interested in how L. pulmonaria hydration air-dried and weighed, and using the methods described above
might be related to ambient moisture sources, we plotted values (see ‘Thallus calibration’, above), these were suspended from
of thallus WC and VPD for different times of the day, grouped the same growth frames as those used for the measurement of
46 Borge and Ellis — Moisture, light and the climate change response of L. pulmonaria

thallus hydration and with the accompanying dataloggers (see 1.0

‘Thallus hydration’, above). At monthly intervals the thalli
were air-dried for 24 h (21 °C, relative humidity ~40 %), and

Weighting for photosynthetic rate

their dry weight measured. The dry mass (DM) growth rate 0.8
(DMgrowth, mg mg−1 d−1) was estimated:
(DWt+1 − DWt ) 0.6
DMgrowth =
DWt
The position of the growth frame was plotted onto a digital 0.4
terrain map at 5-m resolution (OS Terrain, accessed under a
UK Ordnance Survey open data licence). An indicative value
for solar irradiance received by the thalli was then calculated 0.2
per day of the year as the cumulative W h−1 m2, using the Solar
Radiation toolkit (Rich et al., 1994; Fu and Rich, 2002) in
ArcMap v.10 (ESRI, Redlands, CA, USA), and assuming clear 0
sky conditions (proportion of the normal radiation flux that was
diffuse = 0.3, transmitivity = 0.5). Solar irradiance was then dis- 0 1 2 3 4
tributed into hourly values using standard flux equations for a Thallus water content (WC)
solar constant, incorporating: (1) day of the year to calculate
the declination of the sun (radians) and the sun–Earth distance Fig. 2. Statistical model used for weighting the summed photosynthetically ac-
(astronomical units), (2) local time to calculate the hour angle tive radiation (PAR) that drives photosynthesis, by Lobaria pulmonaria thallus
of the sun (radians) and (3) latitude to calculate the solar zenith water content (WC), and therefore reflecting a known response of the photosyn-
thetic rate to variable lichen hydration.
angle (cosine). Values of W h−1 m2 were multiplied by 2.1 to
convert to photosynthetically active radiation (PAR) measured
in µmol m−2 s−1, thereby adjusting the standard conversion factor We supplemented the calculation of 2. (above) with a slightly
of 4.57 to recognize that ~45 % of solar energy is in the PAR modified predictor:
range 400–700 nm.
6. Considering that lichen rates of photosynthesis are maxi-
To develop a functional growth model for L. pulmonaria,
mized when the thallus reaches internal saturation (Solhaug
based on the Iwet hypothesis (Palmqvist and Sundberg, 2000;
et al., 2021), we used the value of WC for saturated L. pul-
Palmqvist et al., 2010), we compared the mean DMgrowth for
monaria thalli presented by Wan & Ellis, (2020), equiva-
each month (mg mg−1 d−1) to the following predictors that were
lent to 1.57, as an optimum value for photosynthesis, with a
calculated for each respective monthly interval:
value of 1. We then fitted a four-parameter log-normal model
1. The proportion of 2-hourly time-points during which the li- (Fig. 2) that declined to zero for WC < 0.2 and declined to
chens were hydrated, with WC ≥ 0.2, and corresponding to 0.5 for the maximum recorded WC of 3.54. The decline for
a thallus water content at which photosynthesis for L. pul- WC values when above the optimum represented an effect
monaria remains active (Gauslaa et al., 2017; Hovind et of suprasaturation as additional water accumulates onto the
al., 2020), and during which the estimated value of PAR thallus surface, thereby reducing photosynthetic rate as gas
(µmol m−2 s−1) ≥ 10, corresponding to the measured light exchange becomes limited (Lange et al., 1996, 2000, 2004).
compensation point for L. pulmonaria under laboratory and We then summed estimated values of PAR, following 2.
field conditions (Sundberg et al., 1997; Pannewitz et al., (above), but having weighted those values by multiplying
2003). This represents the proportion of time during which against the respective value from the log-normal response.
conditions were expected to be conducive to photosynthesis.
We compared the six predictors pairwise using the product-
2. For the above conditions, we summed the estimated values
moment correlation coefficient, before comparing them indi-
of PAR, representing the cumulative available light energy
vidually to monthly DMgrowth using simple linear regression
(µmol m−2 s−1) to drive photosynthesis while the lichens
implemented in R (R Development Core Team 2020).
were hydrated.
3. The proportion of 2-hourly time-points during which the li-
chens were hydrated, with WC ≥ 0.2, and during which the Climate change projections
estimated value of PAR (µmol m−2 s−1) < 10. This represents
the proportion of time during which there might be active To understand how L. pulmonaria growth might respond to
respiration in the absence of photosynthesis. future climate change, and especially the role of hydration and
4. For the above conditions, we summed the measured values photoperiodism, we used model averaging among the linear
of temperature, to represent an observed increase in respi- regressions that had compared the six predictors to DMgrowth
ratory costs at higher temperatures (Sundberg et al., 1997, (see ‘Lichen growth’, above), with Aikaike’s Information
1999). Criterion corrected for small sizes (AICc) to calculate AICc
5. The proportion of 2-hourly time-points during which the weights. These AICc weights enable multiple plausible models
lichens were desiccated, with WC < 0.2, and representing to be used for inference, since they reflect a relative weight
conditions of dormancy with no or limited respiration and of evidence contributed by each model in explaining monthly
photosynthesis. DMgrowth. Accordingly, AICc weights were used to calculate a
 Borge and Ellis — Moisture, light and the climate change response of L. pulmonaria 47

weighted average prediction. The approach was tested through 2.0

comparison with observed values of monthly DMgrowth, before
being applied in a climate change framework.
We sought to compare lichen growth for the period June
2018 to June 2020, referred to here as our experimental period 1.5
(Ep) with growth projected under climate change scenarios. A

Electrical conductance
correction factor was therefore necessary to account for the
way in which climate change scenarios report a difference
in values with respect to standard baselines, referred to here 1.0
as Δscenario/baseline, and considering that Ep was not synchronous
with a standard baseline. First, we extracted UK Met Office
interpolated values of observed instrumental weather data
(Perry and Hollis, 2005; Hollis et al., 2019) for monthly 0.5
mean temperature and relative humidity, and total precipi-
tation, for Ep at a 12-km grid-scale that included the experi-
mental site. Second, we compared these interpolated values
to monthly averages for the standard 30-year climate baseline 0
of 1981–2010, again realized at a 12-km grid-scale including 0 0.5 1.0 1.5 2.0 2.5 3.0
the experimental site, then calculating a difference between Water content (mgH2O per mg dry mass)
the standard baseline, and interpolated climate values for
Ep, referred to as ΔEp/baseline. Third, we adjusted the 2-hourly Fig. 3. Comparison between the water content (WC) of Lobaria pulmonaria
recorded values of temperature and relative humidity, thalli and electrical conductance (direct current, V), showing the mean and
and daily precipitation, by their respective correction as standard error for each of six longitudinal datasets coded with separately col-
Δscenario/baseline minus ΔEp/baseline, providing a set of downscaled oured symbols.
climate change scenarios that accounted for the difference
between the prevailing climatic conditions during the experi-
mental period June 2018 to June 2020 (Ep), and the standard 1.7363
baseline. For the climate change scenarios we adopted the Electrical conductance (y) = x−1.3503

period 2071–2080 and used a seven-member ensemble under 1 + e−[ 0.2566 ]

RCP 8.5, made available by the UKCP18 project (Lowe et Rearranging this equation made it possible to alternatively
al., 2018; Murphy et al., 2018). estimate thallus WC from the measurement of electrical
In summary, our approach synthesized two sets of climate conductance:
data that could be used to calculate VPD and daily precipita- ï ò
tion, those data recorded using dataloggers over the experi- y
WC (x) = 1.3503 + 0.2566 ∗ Ln
mental period at 2-h intervals, and daily precipitation, and a 1.7363 − y
set of respective data that were adjusted according to projected
climate change scenarios. In each case, the previously devel-
oped GLM (see ‘Thallus hydration’, above) was used to predict Thallus hydration
L. pulmonaria hydration above or below a thallus WC ≥ 0.2,
and corresponding to the water content at which photosynthesis Therefore, over a 2-year period from June 2018 to June 2020,
for L. pulmonaria remains active (Gauslaa et al., 2017; Phinney though excluding June 2019 for equipment maintenance, we
et al., 2018), for 100 simulated thalli. Summarized as the hy- were able to reconstruct the real-time hydration of L. pulmon-
dration status at 2-h intervals over each month, and through aria under the experimental field conditions, using the mean
comparison with a static diurnal and seasonal light regime (see WC value of the two thalli that had their electrical conductance
‘Lichen growth’, above), the weighted average prediction of measured at 2-h intervals. Temperature and relative humidity
DMgrowth was estimated and averaged across the 100 simulated were logged for the same period, being used to derive VPD.
thalli, being expressed as the difference in potential growth be- This provided a total of 9238 2-hourly records, over 770 d.
tween the experimental period (Ep) and the projected climate When plotted as a daily trend, grouped per month (Figs 4
change scenario. and 5), there was a clear pattern in which L. pulmonaria hy-
dration responded to VPD. During summer months there was
a pattern of low night-time VPD, with unimodal increase to a
RESULTS high daytime VPD peaking in mid-afternoon, driven by a di-
urnal temperature difference. Thallus WC tracked this change
Thallus calibration in VPD, tending to increase from late afternoon and overnight
to peak in the early morning, before declining during the day
The mean weight ± 1 s.e., for thalli that were used for the cali- to a mid-afternoon minimum. During winter months VPD was
bration of electrical conductance with thallus water content, was much less variable, remaining low throughout the day and night
0.087 ± 0.0036 g. The calibration curve (Fig. 3) was fitted best with consistently high thallus WC. Spring and autumn months
by a three-parameter sigmoidal function (adjusted-R2 = 0.92, tended to have an intermediate pattern, with a weaker diurnal
P < 0.0001 with 78 d.f.): trend in VPD and thallus WC.
48 Borge and Ellis — Moisture, light and the climate change response of L. pulmonaria

June 2018 July 2018 August 2018 September 2018

2.2 1800
2.0 1600
109.4 mm 95.8 mm
1.8
1400
1.6
1200
Mean thallus WC

1.4

Mean VPD
1.2 1000
1.0 800
0.8
600
0.6
176.7 mm 268.4 mm 400
0.4
0.2 200
0 0

October 2018 November 2018 December 2018 January 2019

2.2 1800
2.0 1600
286.8 mm 285.9 mm 371.8 mm 160.5 mm
1.8
1400
1.6
1200
Mean thallus WC

1.4

Mean VPD
1.2 1000
1.0 800
0.8
600
0.6
400
0.4
0.2 200
0 0

February 2019 March 2019 April 2019 May 2019

2.2 1800
2.0 1600
178.7 mm 318.6 mm 61.8 mm 108.8 mm
1.8
1400
1.6
Mean thallus WC

1200
1.4

Mean VPD
1.2 1000
1.0 800
0.8
600
0.6
400
0.4
0.2 200

0 0
06 0
10 0
14 0
18 0
22 0
0

06 0
10 0
14 0
18 0
22 0
0
06 0
10 0
14 0
18 0
22 0
0

06 0
10 0
14 0
18 0
22 0
0
:0
:0

:0
:0
:0
:0
:0
:0

:0
:0
:0
:0
:0
:0
:0
:0
:0
:0
:0
:0

:0
:0
:0
:0

02

02
02

02

Fig. 4. Comparison of mean Lobaria pulmonaria thallus WC (± 1 s.e.), as lines and closed symbols, and mean microclimatic VPD (± 1 s.e.), as dashed lines and
open symbols, calculated at 2-hourly intervals over a daily time-course, for contrasting months of the year; June 2018 to May 2019. Monthly rainfall total in mm.

Exploring the predictors for thallus WC, values of VPD within a probabilistic model (Fig. 6D), for a binomial response
were weakly but negatively correlated with daily rainfall (Fig. threshold of WC ≥ 0.2, corresponding to the thallus water con-
6A; r = −0.246, P < 0.0001), while the relationship between tent at which photosynthesis for L. pulmonaria remains active
thallus WC and daily rainfall (Fig. 6B) appeared weaker than (Gauslaa et al., 2017; Phinney et al., 2018), an optimized GLM
between thallus WC and VPD (Fig. 6C). Framing thallus WC retained both rainfall (χ2 = 23.53, P < 0.0001 with 1 d.f.) and
 Borge and Ellis — Moisture, light and the climate change response of L. pulmonaria 49

July 2019 August 2019 September 2019 October 2019

2.2 1800
2.0 1600
170.7 mm
1.8
1400
1.6
1200
Mean thallus WC

1.4

Mean VPD
1.2 1000
1.0 800
0.8
600
0.6
353.4 mm 253.4 mm 232.3 mm 400
0.4
0.2 200
0 0

November 2019 December 2020 January 2020 February 2020

2.2 1800
2.0 168.8 mm 1600
1.8
1400
1.6
1200
Mean thallus WC

1.4

Mean VPD
1.2 1000
1.0 800
0.8
600
0.6
468.8 mm 366.9 mm 544.5 mm 400
0.4
0.2 200
0 0

March 2020 April 2020 May 2020 June 2020

2.2 1800
2.0 202.5 mm 36.1 mm 121.9 mm 194.1 mm 1600
1.8
1400
1.6
Mean thallus WC

1200
1.4

Mean VPD
1.2 1000
1.0 800
0.8
600
0.6
400
0.4
0.2 200

0 0
06 0
10 0
14 0
18 0
22 0
0

06 0
10 0
14 0
18 0
22 0
0

06 0
10 0
14 0
18 0
22 0
0

06 0
10 0
14 0
18 0
22 0
0
:0
:0
:0
:0
:0
:0

:0
:0
:0
:0
:0
:0

:0
:0
:0
:0
:0
:0

:0
:0
:0
:0
:0
:0
02

02

02

02

Fig. 5. Comparison of mean Lobaria pulmonaria thallus WC (± 1 s.e.), as lines and closed symbols, and mean microclimatic VPD (± 1 s.e.), as dashed lines and
open symbols, calculated at 2-hourly intervals over a daily time-course, for contrasting months of the year; July 2019 to June 2020. Monthly rainfall total in mm.

VPD (χ2 = 2072, P < 0.0001 with 1 d.f.) as fixed effects, for a Lichen growth
statistically significant full model (χ2 = 2212, P < 0.0001 with Focusing on the growth experiment, the six predictors re-
2 d.f.) that explained 28.6 % of the overall deviance. However, lated to the Iwet hypothesis, and that were used to explain
with deviance partitioning, the unique effect of VPD was esti- monthly DMgrowth, displayed contrasting degrees of covariation
mated at 26.8 % and that of rainfall at only 0.3 %. (Table 1). Certain predictors were functionally related as partly
50 Borge and Ellis — Moisture, light and the climate change response of L. pulmonaria

A B
5000 4

4000
3

WC (mgH2O per mg dry mass)

3000
VPD (Pa)

2000

1
1000

0 0

0 20 40 60 80 100 0 20 40 60 80 100

Daily rainfall (mm) Daily rainfall (mm)

C D
4 1.0

0.8
3
WC (mgH2O per mg dry mass)

Probability thallus WC ≥ 0.2

0.6

0.4

1 90th - 22 mm rain
0.2 median - 2 mm rain
10th - 0 mm rain

0 0.0

0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5000

VPD (Pa) VPD (Pa)

Fig. 6. (A) The relationship between the daily rainfall amount and VPD. (B) The relationship between daily rainfall and thallus water content (WC). (C) The re-
lationship between VPD and thallus WC. (D) The modelled response (GLM) for probability of thallus WC ≥ 0.2, compared to VPD while projected for different
daily rainfall amounts, at the 90th, median and 10th percentiles of the observed data.

inverse metrics, for example the proportion of time hydrated these correlated relationships, when tested individually using
in the light compared to the dark (though confounded by the simple linear regression (Fig. 7), DMgrowth was significantly
proportion of time desiccated with WC < 0.2), while other pre- positively related to the proportion of time hydrated in the light,
dictors were related through quantification of a prior categor- the summed PAR while hydrated and the weighted PAR while
ization, for example the cumulative or weighted PAR compared hydrated, while being negatively related to the proportion of
to the proportion of time hydrated in the light. Notwithstanding time hydrated in the dark.
 Borge and Ellis — Moisture, light and the climate change response of L. pulmonaria 51

Table 1. Pairwise correlation of measurements related to the Iwet hypothesis and used to predict DMgrowth for Lobaria pulmonaria ex-
perimental thalli.

1. 2. 3. 4. 5. 6.

1. Time hydrated in light 1

2. Summed PAR while hydrated 0.909*** 1
3. Time hydrated in dark −0.804*** −0.855*** 1
4. Temp. sum while hydrated in dark 0.281 0.345 0.022 1
5. Time desiccated 0.075 0.278 −0.653** −0.394 1
6. Weighted PAR while hydrated 0.901*** 0.959*** −0.745*** 0.53* 0.103 1

***P < 0.001, **P < 0.01, *P < 0.05 for 17 d.f.

Climate change projections hypothesis (Palmqvist and Sundberg, 2000; Palmqvist et al.,
2010). These key requirements were investigated here for the
For subsequent climate change analysis, the proportion of
lichen epiphyte L. pulmonaria.
time hydrated in the light was selected in preference to alter-
First, we addressed the issue of lichen hydration in biocli-
natives that were also related to PAR but had lower adjusted-R2
matic models. A relationship that was used to infer L. pul-
and higher AICc (summed PAR and weighted PAR while hy-
monaria hydration (thallus WC) from electrical conductance
drated). Therefore, representing alternative responses of photo-
had a strongly sigmoidal response (Fig. 3). Previous compar-
synthesis and respiration, the proportion of time hydrated in
able studies (cf. Jonsson et al., 2008) have suggested that this
the light and the proportion of time hydrated in the dark were
sigmoidal pattern may be an artefact of thallus dynamics, if
used for prediction of DMgrowth by multimodel averaging. This
thallus margins (where the clips are placed and electrical con-
led to consensus model predictions that were strongly deter-
ductance is measured) attain a higher WC more rapidly than
mined by the proportion of time hydrated in the light (AICc
central thallus portions during a rehydration phase (Phinney et
weight = 0.93), considering that this is to a degree also redun-
al., 2018; Hovind et al., 2020), and conversely having lower
dant with (and captures) variability in growth explained by the
WC during a desiccation phase. The sigmoidal relationship
proportion of time hydrated in the dark. Consequently, the inde-
may therefore overestimate thallus WC during rehydration and
pendent contribution of the proportion of time hydrated in the
underestimate thallus WC during desiccation. Here we chose to
dark received a lower weight (AICc weight = 0.07). There was
cautiously retain the sigmoidal curve because of the overriding
a positive linear relationship between the observed and model-
strength of statistical fit (being highly non-linear), and because
predicted DMgrowth (Fig. 8A), though with a tendency for the
the functional relevance of thallus moisture dynamics may also
model to underestimate the highest growth rates and to over-
be non-linear, given evidence for higher photosynthetic invest-
estimate the lowest (negative) growth rates. When projected to
ment towards apical and non-reproductive thallus margins (cf.
climate change scenarios (Fig. 8B), the shift in DMgrowth tended
Máguas and Brugnoli, 1996; Tretiach and Carpanelli, 1992;
to be more strongly negative (lower growth) during the summer
Baruffo et al., 2008). Our interpretation of L. pulmonaria hy-
period, with little change for the winter period. The net mean
dration is therefore to be treated cautiously until verified against
shift in annual DMgrowth was negative under the climate change
a framework that integrates moisture dynamics with structural
scenarios.
patterns of photosynthesis, working from the margins towards
the central portion of thallus lobes. Allowing due caution, we
can nevertheless gain some insight into lichen hydration rele-
DISCUSSION vant to bioclimatic models.
There is an imperative to shift climate change analysis from Previous large-scale distributional and bioclimatic studies
a broad measurement of risk based on correlative bioclimatic have frequently used climate indices (Tuhkanen, 1980), or
models (Pearson and Dawson, 2003; Heikkinen et al., 2006) model combinations of standard bioclimatic variables such as
towards an improved functional basis that incorporates physio- temperature averages and precipitation totals as predictors (see
logical response and demographic consequences (Huntley et table 1 in Ellis, 2019a). These might be considered distal cli-
al., 2010; Evans et al., 2015). Achieving this for poikilohydric mate variables (sensu Austin, 2002, 2007); for example, annual
NVPs such as bryophytes and lichens poses two challenges to monthly precipitation total, interpolated at coarse-grained
that are addressed here. First, there is a requirement to incorp- scales (1–10-km grids), is only equivocally relevant to lichen
orate prior understanding that these species are physiologically hydration, physiological response and growth. It may cause a
coupled to hydration achieved by access to multiple sources direct effect by explaining to some uncertain degree hydration
of moisture, as has been demonstrated for lichen epiphytes, at microhabitat scales (Kennedy, 1997), as direct interception
including rain, condensed dew-formation and water vapour but also in relation to stemflow (Van Stan et al., 2014; Cayuela
(Gauslaa, 2014). Second, there is a requirement to incorporate et al., 2018), but may also be a proxy variable that is correl-
prior understanding that the growth of these species is deter- ated with alternative and possibly more relevant features of
mined by a relationship between the period of hydration and the climate such as the number of rain days (Ellis et al., 2017;
light availability to drive photosynthesis, referred to as the Iwet Phinney et al., 2021), or atmospheric humidity (Caldiz, 2004;
52 Borge and Ellis — Moisture, light and the climate change response of L. pulmonaria

A B
0.003
DMgrowth (mg mg–1 d–1)

0.002

0.001

adj-R 2 = 0.44, P = 0.0016; AICc = –200.60 adj-R 2 = 0.39, P = 0.0034; AICc = –198.97
–0.001
0.2 0.3 0.4 0.5 0.6 0.7 0 5.0e+4 1.0e+5 1.5e+5 2.0e+5 2.5e+5

Proportion of time hydrated in light Summed PAR while hydrated

C D
0.003
DMgrowth (mg mg–1 d–1)

0.002

0.001

adj-R 2 = 0.26, P = 0.0185; AICc = –195.44 non-signif.

–0.001
0.2 0.3 0.4 0.5 0.6 0.7 0.8 250 500 750 1000 1250 1500 1750 2000 2250

Proportion of time hydrated in dark Temp. sum while hydrated in dark

E F
0.003
DMgrowth (mg mg–1 d–1)

0.002

0.001

non-signif. adj-R 2 = 0.31, P = 0.0092; AICc = –196.86

–0.001
0 0.1 0.2 0.3 0.4 0.5 0 3.0e+4 6.0e+4 9.0e+4 1.2e+5 1.5e+5

Proportion of time desiccated Weighted PAR while hydrated

Fig. 7. Comparison of Lobaria pulmonaria growth (mean ± 1 s.e.) with each of six predictors related to the Iwet hypothesis; linear regression models are shown
for statistically significant relationships.
 Borge and Ellis — Moisture, light and the climate change response of L. pulmonaria 53

0.0030 0.0001
A B
Predicted DMgrowth (mg mg–1 d–1)

Shift predicted DMgrowth

0.0025
–0.0001
Net mean ± 1SE

0.0020 –0.0002

–0.0003
0.0015
–0.0004

0.0010 –0.0005
–0.001 0 0.001 0.002 0.003 0.004

18

19

19

20

20
01

01

01

01
20

20

20

20

20
r2

r2

r2

r2
Observed DMgrowth (mg mg–1 d–1)

ne

ch

ne

ch

ne
be

be

be

be
ar

ar
Ju

Ju

Ju
em

em

em

em
M

M
pt

ec

pt

ec
Se

Se
D

D
Fig. 8. (A) The relationship between observed Lobaria pulmonaria DMgrowth and values predicted independently based on model projections. (B) The predicted
shift in DMgrowth based on a seven-member ensemble climate change scenario (RCP 8.5, 2070s), each member coded with separately coloured symbols, using a
multimodel combination of photosynthesis (cf. Fig. 7A) and respiration (Fig. 7C), relative to growth during the experimental period.

Merinero et al., 2015) mediated by the cycling of precipita- notwithstanding a need to disentangle these consequences of
tion between the terrestrial surface and evaporative return to VPD variability especially for the hydration of cyanolichens,
the atmosphere (Rowntree and Bolton, 1983; Delworth and where partitioning of liquid and vapour sources needs to be es-
Manabe, 1989; Zhou et al., 2019). Consistent with this cyc- pecially considered (Lange et al., 1986, 1988, 1993). Drawing
ling of precipitation, several studies have related trends in li- on a wider body of evidence, VPD has been used as a predictor
chen diversity to patterns of topographic wetness, including soil to explain lichen growth and distribution (Rambo, 2010; Ellis,
moisture status (Rolstad et al., 2001; Radies et al., 2009; Ellis 2020) and response to environmental change (Rambo and
and Eaton, 2021). Accordingly, the direct effect of standard North, 2012; Song et al., 2014), and has been proposed as a key
bioclimatic variables is often fraught with ecological uncer- metric in explaining both lichen physiological responses and
tainty, and there is a need to develop proximal climate vari- ecosystem feedbacks (Pypker et al., 2016; Stanton et al., 2022).
ables (sensu Austin, 2002, 2007) that can be more confidently Second, we addressed a relationship between the period of
related to lichen hydration. In this regard, the results – gathered hydration and the period of light availability to drive photosyn-
under field conditions of an oceanic high-rainfall environment thesis. The results appear to confirm the importance of seasonal
(Ellis, 2016) – suggest, perhaps surprisingly in this context, that patterns in both climate and light availability when modelling
direct wetting from precipitation is less important as an inde- the lichen response to climate change (Ellis et al., 2017; Ellis,
pendent moisture source for hydration of L. pulmonaria than 2019b). Consistent with the Iwet hypothesis (Palmqvist and
VPD. In support of our observation, a strong link between VPD Sundberg, 2000; Palmqvist et al., 2010), lichen growth was
and lichen hydration has been observed previously (Pypker et positive when a greater proportion of time was spent hydrated in
al., 2016), including for L. pulmonaria (Gaio-Oliveira et al., conditions that exceeded a light compensation point. Allowing
2004); nevertheless, interpretating hydration responses to VPD for the observation that some respiration in the dark can be
is multifaceted. We found that VPD had a significant albeit important for thallus maintenance (Bidussi et al., 2013; Alam
weak relationship with rainfall. To a degree VPD may be cap- et al., 2015), lichen growth was nevertheless negative when
turing the direct effect of rainfall. Furthermore, VPD can be a greater proportion of time was spent hydrated in the dark.
related to the likelihood of condensed dew-formation, as well Similarly, dark-induced respiratory costs have been used to ex-
as to the availability of atmospheric water vapour (Anderson, plain spatial patterns ranging from the lower abundance of fo-
1936; Barry and Blanken, 2016), which is a demonstrated mois- liose lichens comparing the lowland with montane tropics (Zotz
ture source for L. pulmonaria (Phinney et al., 2019; Hovind et and Winter, 1994; Lange et al., 2000; Zotz et al., 2003), and for
al., 2020), and also to the evaporative drying effect of the air L. pulmonaria including growth differences along latitudinal
(Anderson, 1936; Barry and Blanken, 2016). Consequently, gradients (Gaio-Oliveira et al., 2004), or for a given latitude re-
our results indicate that the largest diurnal fluctuations in lated to seasonal differences in the light regime (Larsson et al.,
VPD observed during the summer months, coinciding with 2012). On this basis, previous field and laboratory studies have
the highest values for L. pulmonaria hydration, can probably suggested that the growth of oceanic lichens may be comprom-
be explained by condensed dew-formation, being followed by ised by future drier summers, causing reduced physiological
rapid daytime drying. Furthermore, a consistently stable and activity while there is higher light to drive photosynthesis, com-
low VPD during the winter months would suggest that thalli bined with future wetter winters, causing increased lichen hy-
hydrated by rain or via uptake through high levels of atmos- dration under lower light, leading to respiratory losses (Ellis
pheric water vapour remained more fully hydrated over longer et al., 2017; Ellis, 2019b). These correspond to climate shifts
periods, including during the daytime. Overall, VPD appears to projected for our study region (Lowe et al., 2018; Murphy et
be a promising metric for lichen bioclimatic modelling, though al., 2018). This effect of seasonal climate change and light
54 Borge and Ellis — Moisture, light and the climate change response of L. pulmonaria

availability is supported by our projections of L. pulmonaria

growth to future climate change scenarios, causing a net reduc-
1.0
tion in annual growth. However, the ultimate consequence of
reduced growth rates will depend on the effect on generation

Shift in DMgrowth (respiration)

time (Martínez et al., 2012; Eaton and Ellis, 2014) and subse-
quent interaction with habitat dynamics to exercise control over 0.8
fecundity and population size (Ellis, 2018). The expectation for
a reduced growth rate does indicate risk to the viability of L.
pulmonaria populations under climate change.
Although we demonstrate the potential importance of VPD 0.6
for L. pulmonaria hydration, and a relevance to modelled
growth and climate change response, there are several cav-
eats and advances that can be proposed. First, despite the close
link between VPD and thallus WC observed here, the scale 0.4
of inference in this study is of lower resolution than previous
biophysical models that used different drivers of lichen hydra-
tion, such as the water potential of the air (ψair: Jonsson et al.,
2008; Jonsson Čabrajić et al., 2010). These previous models 0.2
aimed for a parameterization at temporal scales reflecting high- 0.80 0.85 0.90 0.95 1.00 1.05 1.10
resolution processes, such as instantaneous wetting from rain-
Shift in DMgrowth (photosynthesis)
fall or condensed dew-formation, compared to slower hydration
to achieve equilibrium with humid air, with consequences that Fig. 9. Proportional shifts in DMgrowth comparing baseline to climate change
affect photosynthetic activation in real-time (Jonsson Čabrajić scenarios based on a seven-member ensemble climate change scenario (RCP
et al., 2010; Lidén et al., 2010). The approach here is a com- 8.5, 2070s), each member coded with separately coloured symbols, and real-
promise between the real-time response of lichen hydration, ized by an inferred change in photosynthesis compared to respiration; reduction
with its inferred consequences for physiological activity, in growth because of increased respiration appears to be potentially greater than
reduction in growth because of reduced photosynthesis.
balanced against simplicity that aims for a general application
in bioclimatic models. Importantly, the results meet a minimum
requirement to resolve the effects on lichen hydration and
growth of seasonal variation in the diurnal moisture and light become more important as a predictor in driving respiratory
regimes. Second, there remains a need to better parameterize losses.
the functional basis for growth imposed by a balance of photo- Third, there is a need to predict functionally relevant climate
synthesis and respiration, before targeted application in climate variables such as VPD, and light availability, at the micro-
change modelling. This is exemplified by the multimodel aver- habitat scale occupied by lichens in nature (cf. Di Nuzzo et al.,
aging of DMgrowth in relation to the proportion of time spent 2022; Coxson and Coyle, 2003; Nilsson et al., 2022), making
hydrated in the light, compared to the dark. If these predictors it possible to transfer growth projections into natural habi-
(time in the light, time in the dark) were given equal weighting, tats. The role of light regime in driving growth at microhabitat
then the projected net reduction of DMgrowth in response to cli- scales will be sensitive to both the latitude and the ecology
mate change scenarios would be greater (Fig. 9). However, in of the forest or woodland stand. Latitudinally, previous ex-
the consensus model with weights used here, the net reduction amples have contrasted higher summer and lower winter light
is smaller because the proportion of time spent hydrated in the for high-latitude L. pulmonaria sites in Sweden, with lower
light had a lower magnitude effect but received greater AICc summer and higher winter light for lower-latitude L. pulmon-
weight, while the independent effect of the proportion of time aria sites in Portugal (Gaio-Oliveira et al., 2004). Considering
spent hydrated in the dark contributed less weight to the model interaction with the canopy, for the gladed position of this ex-
predictions. This balance of outcomes is probably affected by periment the light availability was referenced against an open
the data structure of what is a comparatively long growth ex- sky light regime, though estimates of the downscaled light re-
periment (~2 years), but which nevertheless offers limited gime beneath the forest/woodland canopy may need to factor
monthly data points for partitioning between confounded pre- tree composition, stem density and phenology, including the
dictors. This caution extends to the absence of a temperature possible indirect effect of climate change on lichen response
effect, since temperature is known to control respiration rates because of shifted tree phenology or composition under cli-
(Sundberg et al., 1997, 1999). However, temperature had a mate change (Chmielewski and Rötzer, 2001; Vitasse et al.,
relatively low diurnal and seasonal amplitude under the experi- 2011). Nevertheless, working towards a functional lichen re-
mental field conditions. For context, during the two winter sea- sponse at microhabitat scales is crucial to inform downscaled
sons combined, mean temperature was 4.68 ± 0.071 °C, with and ecologically accurate predictive bioclimatic models that
a recorded maximum of 15.31 °C and with just 3.4 % of data- can be translated into habitat management to offset climate
logged hours >10 °C. If this pattern starts to change – with hy- change risk. Such studies would also have a transferable value
drated thalli spending longer periods in low winter light but for the improved parameterization and increased resolution of,
at much higher temperatures than those measured and used to for example, carbon-balance models applied to poikilohydric
parameterize the growth models here – then temperature could NVPs (Porada et al., 2013; Nikolić et al., 2023).
 Borge and Ellis — Moisture, light and the climate change response of L. pulmonaria 55

FUNDING Ellis CJ. 2016. Oceanic and temperate rainforest climates and their epiphyte
indicators in Britain. Ecological Indicators 70: 125–133.
This work was funded from within the Scottish Government’s Ellis CJ. 2018. A mechanistic model of climate change risk: growth rates
Rural and Environment Science and Analytical Services and microhabitat specificity for conservation priority woodland epi-
phytes. Perspectives in Plant Ecology, Evolution and Systematics 32:
Division Strategic Research Programme. 38–48.
Ellis CJ. 2019a. Climate change, bioclimatic models and the risk to lichen di-
versity. Diversity 11: 54.
ACKNOWLEDGEMENTS Ellis CJ. 2019b. Interactions of climate and solar irradiance can reverse the
bioclimatic response of poikilohydric species: an experimental test for
We thank Peter Baxter and Sarah Kirton for logistical support Flavoparmelia caperata. The Bryologist 122: 98–110.
in running the lichen experiment at Benmore Botanic Garden. Ellis CJ. 2020. Microclimatic refugia in riparian woodland: a climate change
adaptation strategy. Forest Ecology and Management 462: 118006.
We thank two anonymous reviewers for their comments to im- Ellis CJ, Eaton S. 2021. Climate change refugia: landscape, stand and tree-scale
prove the original manuscript. microclimates in epiphyte community composition. The Lichenologist 53:
135–148.
Ellis CJ, Coppins BJ, Dawson TP. 2007a. Predicted response of the lichen
epiphyte Lecanora populicola to climate change scenarios in a clean-air
LITERATURE CITED region of northern Britain. Biological Conservation 135: 396–404.
Ellis CJ, Coppins BJ, Dawson TP, Seaward MRD. 2007b. Response of
Alam MA, Gauslaa Y, Solhaug KA. 2015. Soluble carbohydrates and relative British lichens to climate change scenarios: trends and uncertainties in
growth rates in chloro-, cyano- and cephalolichens: effects of temperature the projected impact for contrasting biogeographic groups. Biological
and nocturnal hydration. The New Phytologist 208: 750–762. Conservation 140: 217–235.
Allen JL, Lendemer JC. 2016. Climate change impacts on endemic, high- Ellis CJ, Eaton S, Theodoropoulos M, et al. 2015. Epiphyte communities
elevation lichens in a biodiversity hotspot. Biodiversity and Conservation and indicator species. An ecological guide for Scotland’s Woodlands.
25: 555–568. Edinburgh: Royal Botanic Garden Edinburgh.
Anderson DP. 1936. Relative humidity or vapor pressure deficit. Ecology 17: Ellis CJ, Geddes H, McCheyne N, Stansfield A. 2017. Lichen epiphyte re-
277–282. sponse to non-analogue monthly climates: a critique of bioclimatic models.
Austin MP. 2002. Spatial prediction of species distribution: an interface be- Perspectives in Plant Ecology, Evolution and Systematics 25: 45–58.
tween ecological theory and statistical modelling. Ecological Modelling Ettinger AK, Buonaiuto DM, Chamberlain CJ, Morales-Castilla I,
157: 101–118. Wolkovich EM. 2021. Spatial and temporal shifts in photoperiod with
Austin MP. 2007. Species distribution models and ecological theory: a crit- climate change. The New Phytologist 230: 462–474.
ical assessment and some possible new approaches. Ecological Modelling Evans TG, Siamond SE, Kelly MW. 2015. Mechanistic species distribution
200: 1–19. modelling as a link between physiology and conservation. Conservation
Barry RG, Blanken PD. 2016. Microclimate and local climate. Cambridge: Physiology 3: cov056.
Cambridge University Press. Fick SE, Hijmans RJ. 2017. WorldClim 2: new 1-km spatial resolution cli-
Baruffo L, Piccotto M, Tretiach M. 2008. Intrathalline variation of chloro- mate surfaces for global land areas. International Journal of Climatology
phyll a fluorescence emission in the epiphytic lichen Flavoparmelia 37: 4302–4315.
caperata. The Bryologist 111: 455–462. Fordham DA, Mellin C, Russell BD, et al. 2013. Population dynamics can
Bidussi M, Gauslaa Y, Solhaug KA. 2013. Prolonging hydration and active be more important than physiological limits for determining range shifts
metabolism from light periods into nights substantially enhances lichen under climate change. Global Change Biology 19: 3224–3237.
growth. Planta 237: 1359–1366. Frankham R. 1996. Relationship of genetic variation to population size in
Braidwood DW, Ellis CJ. 2012. Bioclimatic equilibrium for lichen distribu- wildlife. Conservation Biology 10: 1500–1508.
tions on disjunct continental landmasses. Botany 90: 1316–1325. Fu P, Rich PM. 2002. A geometric solar radiation model with applications
Caldiz MS. 2004. Seasonal growth pattern in the lichen Pseudocyphellaria in agriculture and forestry. Computers and Electronics in Agriculture 37:
berberina in North-Western Patagonia. The Lichenologist 36: 435–444. 25–35.
Cayuela C, Llorens P, Sánchez-Costa E, Levia DF, Latron J. 2018. Effect of Gaio-Oliveira G, Dahlman L, Máguas C, et al. 2004. Growth in relation to
biotic and abiotic factors on inter- and intra-event variability in stemflow microclimatic conditions and physiological characteristics of four Lobaria
rates in oak and pine stands in a Mediterranean mountain area. Journal of pulmonaria populations in two contrasting habitats. Ecography 27: 13–28.
Hydrology 560: 396–406. Gauslaa Y. 2014. Rain, dew, and humid air as drivers of morphology, function
Chmielewski F-M, Rötzer T. 2001. Response of tree phenology to cli- and spatial distribution in epiphytic lichens. The Lichenologist 46: 1–16.
mate change across Europe. Agricultural and Forest Meteorology 108: Gauslaa Y, Solhaug KA, Longinotti S. 2017. Functional traits prolonging
101–112. photosynthetically active periods in epiphytic lichens during desiccation.
Coxson DS. 1991. Impedance measurement of thallus moisture content in li- Environmental and Experimental Botany 141: 83–91.
chens. The Lichenologist 23: 77–84. Giordani P, Incerti G. 2008. The influence of climate on the distrubution of li-
Coxson DS, Coyle M. 2003. Niche partitioning and photosynthetic response chens: a case study in a borderline area (Liguria, NW Italy). Plant Ecology
of alectorioid lichens from subalpine spruce-fir forest in north-central 195: 257–272.
British Columbia, Canada: the role of canopy microclimate gradients. The Green TGA, Lange OL. 1995. Photosynthesis in poikilohydric plants: a com-
Lichenologist 35: 157–175. parison of lichens and bryophytes. In: Schulze E-D, Caldwell MM. eds.
Crawley MJ. 2013. The R book. Chichester: John Wiley & Sons. Ecophysiology of photosynthesis. Berlin: Springer-Verlag, 319–341.
Delworth T, Manabe S. 1989. The influence of soil wetness on near-surface Green TGA, Nash TH, Lange OL. 2010. Physiological ecology of carbon di-
atmospheric variability. Journal of Climate 2: 1447–1462. oxide exchange. In: Nash TH. ed. Lichen biology. Cambridge: Cambridge
Di Nuzzo L, Benesperi R, Nascimbene J, et al. 2022. Little time left. University Press, 154–183.
Microrefuges may fail in mitigating the effects of climate change. Science Green TGA, Sancho LG, Pintado A. 2011. Ecophysiology of desiccation/re-
of the Total Environment 825: 153943. hydration cycles in mosses and lichens. In: Lüttge U, Beck E, Bartels D.
Eaton S, Ellis CJ. 2014. High demographic rates for the model epiphyte eds. Plant desiccation tolerance. Berlin: Springer-Verlag, 89–120.
Lobaria pulmonaria in an oceanic hazelwood (western Scotland). Fungal Heikkinen RK, Luoto M, Araújo MB, Virkkala R, Thuiller W, Sykes MT.
Ecology 11: 60–70. 2006. Methods and uncertainties in bioclimatic envelope modelling under
Elbert W, Weber B, Burrows S, et al. 2012. Contribution of cryptogamic climate change. Progress in Physical Geography: Earth and Environment
covers to the global cycles of carbon and nitrogen. Nature Geoscience 5: 30: 751–777.
459–462. Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A. 2005. Very high
Ellis CJ. 2013. A risk-based model of climate change threat: hazard, exposure, resolution interpolated climate surfaces for global land areas. International
and vulnerability in the ecology of lichen epiphytes. Botany 91: 1–11. Journal of Climatology 25: 1965–1978.
56 Borge and Ellis — Moisture, light and the climate change response of L. pulmonaria

Hoffman AA, Sgrò CM. 2011. Climate change and evolutionary adaptation. Merinero S, Martínez I, Rubio-Salcedo M, Gauslaa Y. 2015. Epiphytic li-
Nature 470: 479–485. chen growth in Mediterranean forests: effects of proximity to the ground
Hollis D, McCarthy M, Kendon M, Legg T, Simpson I. 2019. HadUK-Grid and reproductive stage. Basic and Applied Ecology 16: 220–230.
- a new UK dataset of gridded climate observations. Geoscience Data Muir PS, Shirazi AM, Patrie J. 1997. Seasonal growth dynamics in the lichen
Journal 6: 151–159. Lobaria pulmonaria. The Bryologist 100: 458–464.
Hovind ABA, Phinney NH, Gauslaa Y. 2020. Functional trade-off hydration Murphy JM, Harris GR, Sexton DMH, et al. 2018. UKCP18 land projec-
strategies in old forest epiphytic cephalolichens. Fungal Biology 124: tions: science report. Exeter: UK Met Office.
903–913. Nascimbene J, Casazza G, Benesperi R, et al. 2016. Climate change fosters
Huntley B, Barnard P, Altwegg R, et al. 2010. Beyond bioclimatic envelopes: the decline of epiphytic Lobaria species in Italy. Biological Conservation
dynamic species’ range and abundance modelling in the context of climate 201: 377–384.
change. Ecography 33: 621–626. Nicotra AB, Atkin OK, Bonser SP, et al. 2010. Plant phenotypic plasticity in
James PW, Hawksworth DH, Rose F. 1977. Lichen communities in the a changing climate. Trends in Plant Science 15: 684–692.
British Isles: a preliminary conspectus. In: Seaward MRD. ed. Lichen Nikolić N, Zotz G, Bader MY. 2023. Modelling the carbon balance in bryo-
ecology. London: Academic Press, 295–413. phytes and lichens: presentation of PoiCarb 1.0, a new model for ex-
Jonsson AV, Moen J, Palmqvist K. 2008. Predicting lichen hydration using plaining distribution patterns and predicting climate-change effects.
biophysical models. Oecologia 156: 259–273. American Journal of Botany 111: e16266.
Jonsson Čabrajić AV, Lidén M, Lundmark T, Ottosson-Löfvenius M, Nilsson AR, Solhaug KA, Gauslaa Y. 2022. The globally rare threatened epi-
Palmqvist K. 2010. Modelling hydration and photosystem II activation in phytic cyanolichen Erioderma pedicellatum depends on a rare combin-
relation to in situ rain and humidity patterns: a tool to compare the perform- ation of habitat factors. The Lichenologist 54: 123–136.
ance of rare and generalist lichens. Plant, Cell & Environment 33: 840–850. Palmqvist K, Sundberg B. 2000. Light use efficiency of dry matter gain in five
Karger DN, Conrad O, Böhner J, et al. 2017. Climatologies at high reso- macro-lichens: relative impact of microclimate conditions and species-
lution for the earth’s land surface. Scientific Data 4: 170122. specific traits. Plant, Cell and Environment 23: 1–14.
Keith DA, Akçakaya HR, Thuiller W, et al. 2008. Predicting extinction risks Palmqvist K, Dahlman L, Jonsson A, et al. 2010. The carbon economy
under climate change: coupling stochastic population models with dy- of lichens. In: Nash TH. ed. Lichen biology. Cambridge: Cambridge
namic bioclimatic habitat models. Biology Letters 4: 560–563. University Press, 184–217.
Kennedy AD. 1997. Bridging the gap between general circulation model Pannewitz S, Schroeter B, Scheidegger C, et al. 2003. Habitat selection
(GCM) output and biological microenvironments. International Journal and light condition: a field study with Lobaria pulmonaria. Bibliotheca
of Biometeorology 40: 119–122. Lichenologica 86: 281–297.
Lange OL, Kilian E, Ziegler H. 1986. Water vapor uptake and photosynthesis Pearson RG, Dawson TP. 2003. Predicting the impacts of climate change on
of lichens: performance differences in species with green and blue-green the distribution of species: are bioclimate envelope models useful? Global
algae as phycobionts. Oecologia 71: 104–110. Ecology and Biogeography 12: 361–371.
Lange OL, Green TGA, Ziegler H. 1988. Water status related photo- Perry M, Hollis D. 2005. The generation of monthly gridded datasets for
synthesis and carbon isotope discrimination in species of the lichen a range of climatic variables over the UK. International Journal of
genus Pseudocyphellaria with green or blue-green photobionts and in Climatology 25: 1041–1054.
photosymbiodemes. Oecologia 75: 494–501. Phinney NH, Solhaug KA, Gauslaa Y. 2018. Rapid resurrection of
Lange OL, Büdel B, Meyer A, Kilian E. 1993. Further evidence that activa- chlorolichens in humid air: specific thallus mass drives rehydration and re-
tion of net photosynthesis by dry cyanobacterial lichens requires liquid activation kinetics. Environmental and Experimental Botany 148: 184–191.
water. The Lichenologist 25: 175–189. Phinney NH, Solhaug KA, Gauslaa Y. 2019. Photobiont-dependent hu-
Lange OL, Green TGA, Reichenberger H, Meyer A. 1996. Photosynthetic midity threshold for chlorolichen photosystem II activation. Planta 250:
depression at high thallus water contents in lichens: concurrent use of gas 2023–2031.
exchange and fluorescence techniques with a cyanobacterial and a green Phinney NH, Gauslaa Y, Palmqvist K, Esseen P-A. 2021. Macroclimate
algal Peltigera species. Botanica Acta 109: 43–50. drives growth of hair lichens in boreal forest canopies. Journal of Ecology
Lange OL, Büdel B, Meyer A, et al. 2000. Lichen carbon gain under tropical 109: 478–490.
conditions: water relations and CO2 exchange of three Leptogium species Phinney NH, Ellis CJ, Asplund J. 2022. Trait-based response of lichens to
of a lower montane rainforest in Panama. Flora 195: 172–190. large-scale patterns of climate and forest availability in Norway. Journal
Lange OL, Büdel B, Meyer A, Zellner H, Zotz G. 2004. Lichen carbon of Biogeography 49: 286–298.
gain under tropical conditions: water relations and CO2 exchange of Porada P, Weber B, Elbert W, Pöschl U, Kleidon A. 2013. Estimating global
Lobariaceae species of a lower montane rainforest in Panama. The carbon uptake by lichens and bryophytes with a process-based model.
Lichenologist 36: 329–342. Biogeosciences 10: 6989–7033.
Larsson P, Solhaug KA, Gauslaa Y. 2012. Seasonal partitioning of growth Porada P, Weber B, Elbert W, Pöschl U, Kleidon A. 2014. Estmating im-
into biomass and area expansion in a cephalolichen and a cyanolichen of pacts of lichens and bryophytes on global biogeochemical cycles. Global
the old forest genus Lobaria. The New Phytologist 194: 991–1000. Biogeochemical Cycles 28: 71–85.
Lidén M, Jonsson Čabrajić AV, Ottosson-Löfvenius M, Palmqvist K, Porada P, Bader MY, Berdugo MB, et al. 2023. A research agenda for
Lundmark T. 2010. Species-specific activation time-lags can explain nonvascular autotrophs under climate change. New Phytologist 237:
habitat restrictions in hydrophilic lichens. Plant, Cell and Environment 1495–1504.
33: 851–862. Pypker TG, Unsworth MH, Van Stan JT, Bond BJ. 2016. The absorption
Lowe JA, Bernie D, Bett P, et al. 2018. UKCP18 science overview report. and evaporation of water vapor by epiphytes in an old-growth Douglas-fir
Exeter: UK Met Office. forest during the seasonal summer dry season: Implications for the canopy
Máguas C, Brugnoli E. 1996. Spatial variation in carbon isotope discrimin- energy budget. Ecohydrology 10: e1801.
ation across thalli of several lichen species. Plant, Cell and Environment R Development Core Team. 2020. R: a language and environment for statis-
19: 437–446. tical computing. Vienna: R Foundation for Statistical Computing.
Martínez I, Flores T, Otálora MAG, et al. 2012. Multiple-scale envir- Radies D, Coxson DS, Johnson C, Konwicki K. 2009. Predicting canopy
onmental modulation of lichen reproduction. Fungal Biology 116: macrolichen diversity and abundance within old-growth inland temperate
1192–1201. rainforests. Forest Ecology and Management 259: 86–97.
Matesanz S, Gianoli E, Valladares F. 2010. Global change and the evolu- Rambo TR. 2010. Habitat preferences of an arboreal forage lichen in a Sierra
tion of phenotypic plasticity in plants. Annals of the New York Academy of Nevada old-growth mixed-conifer forest. Canadian Journal of Forest
Sciences 1206: 35–55. Research 40: 1034–1041.
McCune B, Derr CC, Muir PS, Shirazi A, Sillett SC, Daly WJ. 1996. Lichen Rambo TR, North MP. 2012. Influence of fuel-reduction forest thinning on
pendants for transplant and growth experiments. The Lichenologist 28: growth of an arboreal forage lichen. Forest Ecology and Management 263:
161–169. 208–215.
McCune B, Yang S, Jovan S, Root HT. 2022. Climate and epiphytic Renhorn K-E, Esseen P-A, Palmqvist K, Sundberg B1997. Growth and vi-
macrolichen communities in the four corners region of the U.S.A. The tality of epiphytic lichens. I. Responses to microclimate along a forest
Bryologist 125: 70–90. edge-interior gradient. Oecologia 109: 1–9.
 Borge and Ellis — Moisture, light and the climate change response of L. pulmonaria 57

Rich PM, Hetrick WA, Saving SC, et al. 1994. Using viewshed models to Stanton DE, Ormond A, Koch NM, et al. 2022. Lichen ecophysiology in a
calculate intercepted solar radiation: applications in ecology. American changing climate. American Journal of Botany 110: e16131.
Society for Photogrammetry and Remote Sensing Technical Papers, 524– Sundberg B, Palmqvist K, Esseen P-A, et al. 1997. Growth and vitality of
529. Baton Rouge, LA: ASPRS. epiphytic lichens II. Modelling of carbon gain using field and laboratory
Rolstad J, Gjerde I, Storaunet KO, Rolstad E, Rolstad J. 2001. Epiphytic data. Oecologia 109: 10–18.
lichens in Norwegian coastal spruce forest: historic logging and present Sundberg B, Ekbald A, Näsholm T, et al. 1999. Lichen respiration in rela-
forest structure. Ecological Applications 11: 421–436. tion to active time, temperature, nitrogen and ergosterol concentrations.
Root HT, McCune B, Jovan S. 2014. Lichen communities and species indi- Functional Ecology 13: 119–125.
cate climate thresholds in southeast and south-central Alaska, USA. The Swab RM, Regan HM, Matthies D, et al. 2015. The role of demography,
Bryologist 117: 241–252. intra-species variation, and species distribution models in species’ projec-
Rowntree PR, Bolton JA. 1983. Simulation of the atmospheric response to tions under climate change. Ecography 38: 221–230.
soil moisture anomalies over Europe. Quarterly Journal of the Royal Travis JMJ. 2003. Climate change and habitat destruction: a deadly anthropo-
Meteorological Society 109: 501–526. genic cocktail. Proceedings Biological Sciences 270: 467–473.
Rubio-Salcedo M, Psomas A, Prieto M, Zimmermann NE, Martínez I. Tretiach M, Carpanelli A. 1992. Chlorophyll content and morphology as
2017. Case study of the implications of climate change for lichen diversity factors influencing the photosynthetic rate of Parmelia caperata. The
and distributions. Biodiversity and Conservation 26: 1121–1141. Lichenologist 24: 81–90.
Saikkonen K, Taulavuori K, Hyvönen T, et al. 2012. Climate change-driven Tuhkanen S. 1980. Climatic parameters and indices in plant geography.
species’ range shifts filtered by photoperiodism. Nature Climate Change Uppsala: Acta Phytogeographica Suecica, 67.
2: 239–242. Van Stan JT, Van Stan JH, Levia DF. 2014. Meteorological influences on
Scheidegger C, Frey B, Walser J-C. 1998. Reintroduction and augmenta- stemflow generation across diameter size classes of two morphologically
tion of populations of the endangered Lobaria pulmonaria: methods and distinct deciduous species. International Journal of Biometeorology 58:
concepts. In: Kondratyuk S, Coppins BJ. eds. Lobarion lichens as in- 2059–2069.
dicators of the primeval forests of the Eastern Carpathians. Kiev: N.H. Vitasse Y, François C, Delpierre N, et al. 2011. Assessing the effects of cli-
Kholodny Institute of Botany, 33–52. mate change on the phenology of European temperate trees. Agricultural
Schwartz MW. 1993. Modelling the effects of habitat fragmentation on and Forest Meteorology 151: 969–980.
the ability of trees to respond to climatic warming. Biodiversity and Wan S, Ellis CJ. 2020. Are lichen growth form categories supported by con-
Conservation 2: 51–61. tinuous functional traits: water-holding capacity and specific thallus mass?
Smith GM. 1938a. Cryptogamic botany volume 1. Algae and fungi. London: Edinburgh Journal of Botany 77: 65–76.
McGraw-Hill Book Company. Zhou S, Williams AP, Berg AM, et al. 2019. Land–atmosphere feedbacks
Smith GM. 1938b. Cryptogamic botany volume 2. Bryophytes and pterido- excacerbate concurrent soil drought and atmospheric aridity. Proceedings
phytes. London: McGraw-Hill Book Company. of the National Academy of Sciences 116: 18848–18853.
Smith RJ, Jovan S, McCune B. 2019. Climatic niche limits and community-level Zotz G, Winter K. 1994. Photosynthesis and carbon gain of the lichen,
vulnerability of obligate symbioses. Journal of Biogeography 47: 382–395. Leptogium azureum, in a lowland tropical forest. Flora 189: 179–186.
Solhaug KA, Asplund J, Gauslaa Y. 2021. Apparent electron transport rate - a Zotz G, Schultz S, Rottenberger S. 2003. Are tropical lowlands a marginal
non-invasive proxy of photosynthetic CO2 uptake in lichens. Planta 253: 14. habitat for macrolichens? Evidence from a field study with Parmotrema
Song L, Liu W, Zhang Y, et al. 2014. Assessing the potential impacts of ele- endosulphureum in Panama. Flora - Morphology, Distribution, Functional
vated temperature and CO2 on growth and health of nine non-vascular epi- Ecology of Plants 198: 71–77.
phytes: a manipulation experiment. American Journal of Plant Sciences Zuur AF, Ieno EN, Walker NJ, et al. 2009. Mixed effects models and exten-
05: 1587–1598. sions in ecology with R. New York: Springer.