Ta

WLa

b

c

d

e

f

a

ARRA

KCEMREA

1

oaPa

0h

Agricultural and Forest Meteorology 189–190 (2014) 220–228

Contents lists available at ScienceDirect

Agricultural and Forest Meteorology

j o ur na l ho me pag e: www.elsev ier .com/ locate /agr formet

iming and duration of phenological sequences of alpine plants alongn elevation gradient on the Tibetan plateau

ang Shipinga,∗, Wang Changshuna,e, Duan Jichuangb,c, Zhu Xiaoxueb, Xu Guangpingd,uo Caiyunb, Zhang Zhenhuab, Meng Fandonga,e, Li Yingnianb, Du Mingyuanf

Laboratory of Alpine Ecology and Biodiversity, Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing 100101, ChinaKey Laboratory of Adaptation and Evolution of Plateau Biota, Northwest Institute of Plateau Biology, Chinese Academy of Sciences, Xining 810008, ChinaBinhai Research Institute in Tianjin, Tianjin 300457, ChinaGuangxi Institute of Botany, Guangxi Zhuang Autonomous Region and Chinese Academy of Sciences, Guangxi 541006, ChinaBiological Department, Graduate University of Chinese Academy of Sciences, Beijing 100049, ChinaNational Institute for Agro-Environment Science, Tsukuba 305-8604, Japan

r t i c l e i n f o

rticle history:eceived 11 October 2013eceived in revised form 23 January 2014ccepted 27 January 2014

eywords:limate changearly-spring flowering plantsid-summer flowering plants,

eproductive phenologyvolutionary adaptationlpine plants

a b s t r a c t

Previous studies have focused on the effects of increased temperatures on a single green-up and/orflowering event, but less is known about how acceleration of spring phenology may change subsequentphenological events. We present results of a field experiment to test the hypotheses that (1) the timingof phenological events does not necessarily delay as elevation increases; (2) changes in the timing ofa sequence of phenological events will be consistent for all phenological events along the elevationgradient; and thus (3) change in the timing of phenological events does not affect the duration of theentire reproductive stage in the alpine region. The experiment was conducted along an elevation gradientfrom 3200 to 3800 m using two early-spring flowering (ESF) sedges and four mid-summer flowering(MSF) plants (two forbs and two grasses). Generally, our results only supported the first hypothesis.Lower elevation delayed the starting dates of all phenological events for ESF plants at 3200 m comparedwith other elevations, whereas the opposite trend was observed for MSF-grasses. MSF-forbs had theearliest leaf-out at 3200 m and the earliest first flowering at 3600 m, and onset of fruit-set advancedwith increasing elevation. The entire reproductive duration was shortened with increasing elevation forMSF-forbs, whereas it was the shortest for ESF at 3600 m and for MSF-grasses at 3200 and/or 3800 m.Individual reproductive stages had independent responses to climate change. The duration of the entiregrowing season for ESF plants decreased as elevation increased. For MSF-forbs, it was longest at 3200 mand shortest at 3400 m, while for MSF-grasses it was shortest at 3200 m and at 3800 m. Reproduction was

compressed into shorter time periods only for MSF-forbs at 3600 and 3800 m. Therefore, reproductionis not tightly integrated across the life cycle, and earlier reproductive development induced by warmerspring temperatures did not consistently advance flowering and fruiting times and their durations forthe alpine plants studied. The effects of climate change on the timing and duration of phenological eventswere species-specific. Selection for changes in the timing and duration of individual phenological stagesin response to climate change due to evolutionary adaptation should be taken into account.

. Introduction

Previous phenological studies have focused on the effectsf increased temperatures on a single phenological event such

s flowering through observational studies (Miller-Rushing andrimack, 2008; Dunnell and Travers, 2011; Wolkovich et al., 2012)nd climate manipulation experiments (Arft et al., 1999; Dunne

∗ Corresponding author. Tel.: +86 10 84097096.E-mail address: wangsp@itpcas.ac.cn (S. Wang).

168-1923/$ – see front matter © 2014 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.agrformet.2014.01.021

© 2014 Elsevier B.V. All rights reserved.

et al., 2003; Walker et al., 2006; Sherry et al., 2007, 2011), andby interpreting the onset and end of the growing season throughremote sensing (Yu et al., 2010; Shen et al., 2011; Piao et al., 2011).Less is known about how accelerating green-up and/or floweringmay change subsequent phenological events (Post et al., 2008).Recently, the effects of climate change on seasonal phenologicalevents (i.e., leaf unfolding, flowering, fruit and coloring) for onespecies relative to those of an interacting species have become

increasingly apparent (Durant et al., 2005; Visser and Both, 2005;Post et al., 2008; Yang and Rudolf, 2010). Therefore, it is impor-tant to understand how sequences of phenological events respondto climate change. In particular, understanding both the timing

st Me

atre

tchpme2rtaas

otp2osmnsrtetvnnlf(d

2

2

EoAe23

S. Wang et al. / Agricultural and Fore

nd duration of each phenological event will give a more informa-ive picture of where the greatest or least flexibility in life historyesponse to warming occurs in the life cycle of organisms (Postt al., 2008; Sherry et al., 2011).

Alpine ecosystems on the Tibetan plateau are very sensitiveo climate change (IPCC, 2007). In order to better predict climatehange impacts on those ecosystems, scientists have been assessingow climate change is affecting – and will affect in the future –lant phenology at the regional scale using remote sensing and/orodel derived information as proxies for measurement of veg-

tation green-up (Yu et al., 2010; Shen et al., 2011; Piao et al.,011). Few manipulative experiments have been performed in theegion (Dorji et al., 2013). Moreover, almost no data is available onhe effects of sequences of phenological events on the response oflpine plants to climate change. Here we intend to fill this gap using

3-year field experiment along an elevation gradient involvingites at 3200, 3400, 3600 and 3800 m on the Tibetan plateau.

Although elevation drives phenological phases mainly by meansf a temperature decrease of about 0.6 ◦C every 100 m (Barry, 1981),he cumulative heat requirement (i.e., growing degree-days) ofhenological events also decreases as altitude increases (Shen et al.,011). Organisms may experience trade-offs between adjustmentf one phenological event to climatic change and the timing of sub-equent phenological events. Such trade-offs may determine theanner in which both the timing and duration of successive phe-

ological events respond to climate change (Post et al., 2008). Sometudies show that the effects of climate change on the timing ofeproduction is greater than on the duration of reproduction, andhat responses are species-specific (Price and Waser, 1998; Postt al., 2008). We test three hypotheses in our study, namely that (1)he timing of phenological events does not necessarily delay as ele-ation increases, because this depends on the balance of the mag-itudes of changes in temperature and heat requirements for phe-ological events; (2) changes in the timing of a sequence of pheno-

ogical events will be consistent (i.e., earlier–earlier or later–later)or all phenological events along the elevation gradient; and thus3) change in the timing of phenological events does not affect theuration of the entire reproductive stage in the alpine region.

. Materials and methods

.1. Experimental design

The experiment was conducted at Haibei Alpine Meadowcosystem Research station (HBAMERS) of the Chinese Academyf Sciences, located at latitude 37◦37′ N, and longitude 101◦12′ E.

southern slope of the Qilian Mountains from 3200 to 3800 m inlevation was selected for the elevation gradient (Fig. 1) and four0 m long × 10 m wide sites were fenced at 3200, 3400, 3600 and800 m in 2006 Three plots of 1 m × 1 m were identified, and six

Fig. 1. Diagram of the landscape and experimental sites. ( ) Experimental sites.

teorology 189–190 (2014) 220–228 221

common plant species from these plots were chosen for monitor-ing of all phenological events at each elevation during the growingseasons of 2008–2010. The plant species observed were two early-spring flowering (ESF) species (i.e., Kobresia humilis (Kh) and Carexscabrirostris (Cs)) which flower before May, and four mid-summerflowering (MSF) species (i.e., two grasses, Poa pratensis (Pp) andStipa aliena (Sa), and two forbs, Potentilla anserine (Pa) and Potentillanivea (Pn)), which flower between late June and July.

2.2. Air temperature and soil moisture

At the center of each site, HOBO weather stations (Onset Com-puter Corporation, Cape Cod, Massachusetts, USA) were used tomonitor air temperature and soil moisture (SM) at 20 cm soil depth.Model S-THB-M002 air temperature/RH smart sensors (±0.2 ◦C;±25% RH) with Model RS3 louvered naturally ventilated solar radi-ation shields were used to ensure high accuracy measurements.Model S-SMC-M005 ECH2O soil moisture smart sensors (±3%) wereinstalled horizontally into undisturbed soil by digging a hole andpushing the probes into the side of the hole. Data were sampled at 1-min intervals, and 30-min averages were stored in the data logger.

2.3. Measurement of phenological events

Observations were made at an interval of three or four days fromearly April to the end of October in each year and recordings weremade of the onset dates of seven phenological events, includingemergence of first leaf (EFL), bud/boot-set (BS); first flowering (FF),first fruit-set for forbs or seeding-set for graminoids (FFS), vege-tative stage after fruit/seeding (VAFS), first leaf-coloring (FFL) andthe date of complete leaf-coloring (ELC) (Table 1). Ten individualsfor forbs and ten stems for graminoids of each plant species in eachplot along the elevation gradient were marked during the previ-ous autumn so individual plants could be followed throughout thegrowing season. The first date of each phenological event presenton each marked plant was noted and an unweighted average calcu-lated for each species, i.e., the first day of each phenological eventwas calculated as the day of the year on which phenological charac-teristics were visible for 10% of individuals or 10% of marked stemsof a species. The end date of leaf-coloring was calculated as the dayof the year on which 90% of leaves were colored on the markedindividuals or stems. If a marked plant died or was grazed duringobservations, another nearby stem was marked to replace it, whenavailable. The duration of each phenological event was calculatedas the average number of days between successive phenologicalevents for all individuals or stems in a species. The entire repro-ductive stage included three observed developmental phenologicalevents (i.e., a bud event, a flowering event, and a fruiting event),but excluded the event of seed ripeness because this was difficultto monitor in the field. No data were obtained at 3600 m in 2010because mice destroyed the plots.

2.4. Data analysis

A univariate general linear model (GLM) was applied for anal-ysis of variance using SPSS 160 version. Fixed factors were year,plant species and elevation, and the dependent variables were thetiming and duration of all phenological events measured. One-way ANOVA on timing and duration of each phenological eventwas analyzed along the elevation gradient. Multi-comparison ofleast significant difference (LSD) was conducted for all measuredvariables. Simple correlation analysis was performed between the

timing and duration of all phenological events and to test the pos-sible dependency of the timing and duration of all phenologicalevents on annual/monthly air temperature and soil moisture. Sig-nificant difference is shown at 0.05 level in the text.

222 S. Wang et al. / Agricultural and Forest Meteorology 189–190 (2014) 220–228

Table 1Definition of phenological stages and threshold used to define start of event for the two groups of plants studied.

Phenological stage Phenologicalevent

Forbs Graminoids Threshold (% of markedplants or stems)

Vegetative EFL Date of first visible leaf Date of first visible leaf 10Reproduction BS Date of first unopened bud Date of first spikelet presence (out of boot) 10

FF Date of first open flower Date of first exerted anther or style 10FFS Date of initiation of fruiting Date of passing of the presence of anthers and

styles (developing fruit, separated glume tips)10

Vegetative after fruiting VAFS Date of end of fruiting Date of disarticulating florets 90Leaf-coloring FLC First leaf-coloring First leaf-coloring 10

CLC Complete leaf-coloring Complete leaf-coloring 90

E for fol incluP n are

3

3

−rt2ra

FL, emergence of first leaf; BS, bud/boot-set; FF, first flowering; FFS, first fruit-set

eaf-coloring and CLC, complete leaf-coloring. The 6 alpine plant species measuredp, Poa pratensis and Sa, Stipa aliena. Kh, Cs, Pp and Sa are graminoids, and Pa and P

. Results

.1. Air temperature and soil moisture

Annual average air temperatures over the 3-year period were0.24, −101, −100 and −182 ◦C at 3200, 3400, 3600 and 3800 m,

espectively, and thus decreased by about 0.35 ◦C per 100 m eleva-

ion along the gradient (Fig. 2A). Annual average soil moisture at0 cm was 265, 186, 266 and 82% at 3200, 3400, 3600 and 3800 m,espectively. During the growing seasons from April to October,verage soil moisture was 352, 258, 40.5 and 150% at 3200, 3400,

Fig. 2. Monthly average air temperature and soil m

rbs or seed-set for graminoids; VAFS, vegetative stage after fruit/seeding; FLC, firstde: Kh, Kobresia humilis; Cs, Carex scabrirostris; Pa, Potentilla anserine; Pn, P. nivea;forbs.

3600 and 3800 m, respectively, and during the non-growing sea-sons was 179, 114, 127 and 15% at these same elevations. Growingseason mean soil moisture at 20 cm was highest at 3600 m duringthe growing season, particularly from June to September in 2009and 2010 (Fig. 2B).

3.2. Timing of phenological events

The starting dates of all phenological events were affected byyear, species, elevation and their interactions (p < 0.001) (Table S1).For ESF plants (i.e., Kh and Cs), starting time of leaf-out (EFL) and

oisture at 20 cm depth from 2008 to 2010.

S. Wang et al. / Agricultural and Forest Meteorology 189–190 (2014) 220–228 223

F L, emef . Mean0 (C); Pn

fievwoM

ig. 3. Date of phenological phases for alpine plants along the elevation gradient. EFor graminoids; VAFS, vegetative stage after fruit/seeding; CFL, coloring of first leaf.05 level. Kh, Kobresia humilis (A); Cs, Carex scabrirostris (B); Pa, Potentilla anserine

rst leaf-coloring (FLC) significantly advanced with an increase inlevation, whereas other phenological events did not consistently

ary with elevation, although they were latest at 3200 m comparedith other elevations (Fig. 3A and B). There were different patterns

f variation in phenological events along the elevation gradient forSF plants (i.e., forbs Pa and Pn and grasses Pp and Sa) (Fig. 3C–F).

rgence of first leaf; BS, bud/boot-set; FF, first flowering; FS, fruit for forbs or seedings ± se are shown in the Figures. Different letters indicate significant differences at, P. nivea (D); Pp, Poa pratensis (E) and Sa, Stipa aliena (F).

For Pa and Pn, EFL was earliest and complete leaf-coloring (CLC) waslatest at 3200 m, but most other phenological events were earliest

at 3600 m, and the starting and end dates of FFS and EFS were latestat 3200 m (Fig. 3C and D). For Pp and Sa, most phenological eventswere delayed with an increase in elevation, except for the startingdate of leaf-coloring (FLC) (Fig. 3E and F).

224 S. Wang et al. / Agricultural and Forest Meteorology 189–190 (2014) 220–228

F nt. VSV in theh Poa p

3

sbteawssdw3

ig. 4. Duration of phenological phases for alpine plants along the elevation gradieAFS, vegetative stage after fruit-set; SS, senescence stage. Means ± se are shown

umilis (A); Cs, Carex scabrirostris (B); Pa, Potentilla anserine (C); Pn, P. nivea (D); Pp,

.3. Duration of phenological events

The durations of all phenological events were affected by year,pecies, elevation and their interactions (p < 0.001) (Table S2). Foroth ESF and MSF plants, there were no consistent variations inhe duration of each phenological event with elevation (Fig. 4). Forxample, for ESF plants (i.e. Kh and Cs), the durations of bud-set (BE)nd fruit-set (FSE) were shortest but the durations of flowering (FE)ere longest at 3800 m, whereas the durations of the vegetative

tage before bud-set (VE) were longer and the durations of FE were

horter at 3200 m (Fig. 4A and B). For MSF-forbs (i.e., Pa and Pn),urations of VE were longest at 3200 m and shortest at 3600 m,hile the duration of flowering (FE) was longer at 3200 m than at

600 m for Pn (Fig. 4C and D). For MSF-grasses (i.e., Pp and Sa), the

, vegetative stage; BS, budding stage; FS, flowering stage; FSS, fruit/seeding stage; figures. Different letters indicate significant differences at 0.05 level. Kh, Kobresiaratensis (E) and Sa, Stipa aliena (F).

durations of BE were shortest but durations of FE and FSE werelongest at 3800 m. However, at 3200 m, although the durations ofBE and FSE were longest, the durations of FE were shortest for Ppand Sa (Fig. 4E and F).

3.4. Durations of entire reproduction and growing season

The entire reproductive duration was longest at 3600 m for ESFplants (i.e. Kh and Cs) (Fig. 5). There were no significant differences

in the entire reproductive duration for Cs at 3200, 3400 and 3800 m,whereas for Kh it was longest at 3200 m and shortest at 3400 m(Fig. 5A). For Pa and Pn, the entire reproductive duration decreasedas elevation increased, but for Pp and Sa was longest at 3800 m,

S. Wang et al. / Agricultural and Forest Me

Fig. 5. Variation with elevation of the duration of reproductive stage from bud-ding/boot event to end of fruit/seeding event (A and B) and the ratio of the duration ofreproduction and growing season (C) for different alpine plants. Kh, Kobresia humilis;Cas

w3

iiaadssa

3e

t

s, Carex scabrirostris; Pa, Potentilla anserine; Pn, Potentilla nivea; Pp, Poa pratensisnd Sa, Stipa aliena. Means ± se are shown in the Figures. Different letters indicateignificant differences at 0.05 levels.

hile there was no significant difference for Pp between 3200 and800 m (Fig. 5A).

The duration of the growing season decreased as elevationncreased for ESF plants (i.e., Kh and Cs) (Fig. 5B). For Pa and Pn,t was longest at 3200 m but shortest at 3400 m, whereas for Ppnd Sa it was longest at 3400 and 3600 m, and for Sa was shortestt 3800 m (Fig. 5B). Variations with elevation in the ratios of theuration of reproduction and the growing season were almost con-istent with the patterns of the duration of the entire reproductivetage for these alpine plants, except for Kh between 3200 and 3800nd for Pa between 3200 and 3400 m (Fig. 5C).

.5. Relationships between timing and duration of phenological

vents

Generally, there were significantly positive correlations amonghe phenological events (Table S3), suggesting that earlier leaf-out

teorology 189–190 (2014) 220–228 225

was associated with earlier budding, flowering, fruiting and color-ing. The earlier the starting date of leaf-out, bud-set and flowering,the longer the duration of blooming. Correlations between the starttimings of budding, flowering, fruiting and the end of complete leaf-coloring and the durations of entire reproduction were significantand positive, and the duration of the growing season was mainlyaffected by the start and end timing of leaf-coloring rather thanthe start timing of leaf-out. The longer the vegetative stage beforebudding, the shorter the duration of blooming, but the longer theduration of fruit-set (Table S3).

3.6. Factors affecting the timing and duration of phenologicalevents

3.6.1. Start timing of phenological eventsThere were significantly positive correlations between annual

average air temperature and the start timings of all phenologicalevents and the end timing of leaf-coloring only for ESF plants (i.e.,Kh and Cs) (Table S4), and between annual average soil moistureand the timing of flowering for Kh and the timings of leaf-coloringfor Kh and Cs (Table S5). There were negative correlations werefound between annual average soil moisture and the timings of leaf-out for MSF plants and bud-set for Pn, Pp and Sa, and the timingsof flowering and/or fruit-set for Pp and Sa (Table S5).

Negative correlations for Kh, Cs, Pa and Pn, but positive cor-relations or insignificant correlations for Pp and Sa, were foundbetween monthly air temperature in November and December ofthe previous year and the start timings of all phenological events(Table S4). Positive correlations for Kh and Cs and negative correla-tions for Pp and Sa were observed between monthly soil moisture inNovember and December of the previous year and the start timingsof all phenological events except for leaf-coloring (Table S5).

There were negative correlations between monthly air temper-ature from March to July and the start timings of phenologicalevents except leaf-coloring for Pp and Sa, and between monthlyair temperature from March to May/or June and the start timingsof leaf-out and/or bud-set for Pa and Pn. However, positive correla-tions were found between monthly air temperature in March andMay and the start timings of flowering and/or leaf-coloring for Khand Cs, and between monthly air temperature in June and July andthe start timings of flowering, fruit-set and leaf-coloring for Pa andPn (Table S4). Positive correlations were found between monthlysoil moisture from January to May and the start timings of flower-ing for Kh and the timings of leaf-coloring for Kh and Cs, whereassome negative correlations were observed between monthly soilmoisture from January to June/July and the start timings of leaf-out,bud-set, flowering and fruit-set for Pa, Pn, Pp and Sa (Table S5).

3.6.2. Duration of phenological eventsThere were positive correlations between annual average air

temperature/annual average soil moisture and the duration of thevegetative stage before bud-set for Kh, Sa and/or Pa, of bud-set forKh, Pn, Pp and Sa, of flowering for Pn, and of fruit-set for Kh, Csand/or for all plants except Pp. However, negative correlations werefound between them and the duration of flowering for Kh, Cs, Ppand Sa (Tables S6 and S7), and positive correlations for Pa and Pnwere found between annual average soil moisture and the dura-tions of flowering and of the growing season for all plants exceptPp (Table S7).

There were negative correlations between monthly air tem-perature and soil moisture in November and/or December of theprevious year and the durations of the vegetative stage before bud-

ding and of the growing season for all plants and the duration offruit-set for Pa, Pn and/or Kh and Sa (Tables S6 and S7). Positive cor-relations were found between them and the duration of floweringfor Kh, Cs and Sa and the duration of the growing season for Kh, Cs

2 st Me

aif(

afdmgbcsTbdwnatm

4

4

nbtE(btwecTttes(eitlrftetWttstSh(wnm(m

26 S. Wang et al. / Agricultural and Fore

nd Pa (Tables S6 and S7), and between monthly air temperaturen November and/or December of the previous year the duration ofruit-set for Sa, and the duration of leaf-coloring for Cs, Pa and PnTable S6).

There were positive correlations between monthly air temper-ture from March to June and/or July and the duration of bud-setor all plants, and the duration of fruit-set for Pa and Pn and theuration of the growing season for Kh, Cs, Pa and Pn, and betweenonthly air temperature in April and/or May and the duration of the

rowing season for Pp and Sa, but there was a negative correlationetween it and the duration of fruit-set for Sa (Table S6). Positiveorrelations for Pa and Pn were found between monthly averageoil moisture before July and the duration of flowering (Table S7).here were positive correlations between monthly soil moistureefore June and the duration of bud-set for all alpine plants. Theuration of fruit-set for Kh, Sa, Pa and Pn was positively correlatedith monthly soil moisture at 20 cm depth in April and May, butegative correlations were found between monthly soil moisturend the duration of fruit-set for Cs. There were positive correla-ions between the duration of the growing season and monthly soil

oisture before June and/or July for all plant species (Table S7).

. Discussion

.1. Effects of climate change on timing of phenological events

Our results support the first hypothesis (i.e., that timing of phe-ological events does not necessarily delay as elevation increases),ut do not support the second hypothesis (i.e., that changes in theiming of phenological events is consistent as elevation increases).arlier flowering plants have been shown to mature fruit earlierPenuelas et al., 2002; Post et al., 2008), especially in species thatloom early in the growing season (Sherry et al., 2007). We foundhat the correlations among the start timings of phenological eventsere significantly positive (Table S3), but that in general change in

levation did not consistently influence phenological sequences byonsistently advancing or delaying the timing of all events (Fig. 3).he magnitude of responses of the different phenological eventso climate change varied with elevation and plant species or func-ional group (i.e., ESF vs. MSF plants, or graminoids vs. forbs). Forxample, most of the start timings of phenological events were con-istently later for ESF (i.e., Kh and Cs) and earlier for MSF-grassesi.e., Pp and Sa) at 3200 m, whereas the start timing of leaf-out wasarliest for MSF-forbs (i.e., Pa and Pn) at 3200 m, while start tim-ngs of other phenological events were later. In particular, all ofhe events were earliest at 3600 m except for the start timing ofeaf-out for Pa and Pn. These findings may be attributed to threeeasons. First, several studies have shown that the most importantactors for plant development in alpine areas are temperature andhe date of snowmelt (Walker et al., 2006; Forrest et al., 2010; Dorjit al., 2013), and that late-flowering species are less responsive tohe timing of snowmelt than species that flower early (Price and

aser, 1998). However, some studies suggest that the timing of soilhaw and subsequent soil water availability, rather than snowmeltiming, are the main environmental cues for plant phenology inome arid, semi-arid or monsoon-dominated ecosystems, such ashe Tibetan plateau (Archibald and Scholes, 2007; Wang et al., 2008;hen, 2011; Dorji et al., 2013). We found that at lower elevation,igher temperature with earlier snowmelt delayed spring eventsi.e., leaf-out, bud-set and first flowering date) for ESF. However,ith the exception of the first flowering date for Kh, there were

o significant correlations between spring events and monthly soiloisture from January to July, and we even found that for ESF

i.e., Kh and Cs) positive correlations may indicate that higher soiloisture in November and December of the previous year induces

teorology 189–190 (2014) 220–228

later spring events. In contrast, higher soil moisture and/or highertemperature in November and December of the previous year maycause earlier spring events for ESF and especially for MSF-forbs (i.e.,Pa and Pn) at 3600 m in our study. Similarly, both advances anddelays were observed for the effects of increasing temperature onfruit-set (Menzel et al., 2003). We also found that warming duringthe growing season advanced the fruit-set date for Pp and Sa butdelayed it for other plants, while the opposite effects of warmingin winter were observed. Previous studies indicate that warmingdelays the timing of leaf-coloring in tundra areas (Arft et al., 1999;Walker et al., 2006) and in Europe (Walther et al., 2002). However,we found that there were different patterns of variation with ele-vation for the alpine plants (Fig. 3). Moreover, higher temperaturein June and July may induce earlier leaf-coloring for MSF-grasses(i.e., Pp and Sa), but later leaf-coloring for Kh, Cs and Pn.

Second, most perennial ESF plants develop their floral pri-mordium in the prior autumn and winter, whereas for MSF plantsfloral primordium differentiation is synchronized with vegetativegrowth before flowering (Körner, 1999). Thus, the effects of tem-perature and soil moisture on spring events may be greater forESF plants than for MSF plants in the alpine region. Prior stud-ies documented that responses to spring temperature have beenstronger in early-flowering species that in species that normallyflower later in summer (Menzel et al., 2006; Willis et al., 2010;Cook et al., 2012). Spring flowering phenology is often found todepend most closely on temperatures in the preceding autumn orwinter (Fitter et al., 1995), especially for species that preform springflower buds in the late summer or autumn of the preceding year.We also found that timings of spring events were more stronglycorrelated for ESF plants with mean monthly temperature and soilmoisture in November and December of the previous year than forMSF plants, suggesting that timings of spring events for ESF plantswere better predicted by and more responsive to mean monthlytemperature and soil moisture during winter than for MSF plants.However, our results showed that vernalization–chilling require-ments may not be an important pathway influencing plant responseto spring warming, because there were negative or no correlationsbetween temperature during the previous winter and spring eventsfor these plants. Thus, spring cooling (Piao et al., 2011; Shen, 2011)rather than winter warming (Yu et al., 2010) may have resulted inthe trend of delayed green-up over the Tibetan plateau during the2000s.

Third, although studies indicate that flowering phenology is typ-ically earlier for plants at lower elevations than for plants of thesame species that grow at higher elevations (Bertiller et al., 1990;Ziello et al., 2009), we did not find this phenomenon for Kh, Cs, Paand Pn in the alpine region. Some warming experiments indicatedthat warming reduces soil moisture, which may dampen earlierflowering of plants (Wolkovich et al., 2012). However, we foundthat occurrence of higher temperature and higher soil moisturewas more closely synchronized at lower elevations compared withhigher elevations, except at 3600 m after June. Flowering time isfrequently under selection due to a combination of abiotic, bioticand intrinsic factors (Fitter and Fitter, 2002; Menzel et al., 2005;Sherry et al., 2007; Miller-Rushing and Primack, 2008; Chuine et al.,2010). Evolution in response to this selection is likely to alter notonly flowering time but also reproductive phenology and, poten-tially, traits throughout the life cycle (Galloway and Burgess, 2012).Abundant evidence shows that natural selection can cause evolu-tionary change (Franks et al., 2007; Skelly et al., 2007; Schleip et al.,2008; Colautti et al., 2010; Hermanand and Sultan, 2011; Gallowayand Burgess, 2012; Woods et al., 2012), which may compensate or

even offset the effects of climate change. We found that tempera-ture and soil moisture were greater at lower elevations comparedwith higher elevations, for these alpine plants growing-degree days(GDD) before flowering were also greater at lower elevations, and

S. Wang et al. / Agricultural and Forest Me

Fig. 6. Growing-degree days (GDD) requirement prior to flowering for differentplants at different elevations. Kh, Kobresia humilis; Cs, Carex scabrirostris; Pa, Poten-ti

GEt(eltaplwbap2

4p

cdcsttabaelcfsotrwardaaOt

illa anserine; Pn, P. nivea; Pp, Poa pratensis and Sa, Stipa aliena. Means ± se are shownn the figure.

DD prior to flowering decreased with increasing elevation (Fig. 6).specially for Pa and Pn, the small GDD with higher soil mois-ure prior to flowering at 3600 m may cause their earlier floweringFig. 3). Therefore, compared with the plant species grown in lowerlevations, for the same plant species grown at higher elevationsower basic GDD before flowering partially compensates for loweremperature at higher elevations, which may alter the trend of vari-tion in flowering phenology with elevation. In particular, for ESFlant species (i.e., Kh and Cs), compared with higher elevations,

ater flowering at 3200 m can enable plants to avoid frost damagehere there is a higher risk of frost damage with earlier snowmelt,

ecause the likelihood and degree of frost damage to flower budsre strongly affected by the date of snowmelt (Inouye, 2008) andlants are usually frost damaged at snow-free sites (Hacker et al.,011).

.2. Effects of climate change on duration of the reproductiveeriod and growing season

Our results support the third hypothesis in our study (i.e., thathange in the timing of phenological events does not affect theuration of the entire reproductive stage). Previous studies indi-ated that warming prolonged the duration of the plant growingeason (Arft et al., 1999; Walker et al., 2006). However, we foundhat variation in the duration of the growing season with eleva-ion depended on plant species (Fig. 4). For ESF plants (i.e., Khnd Cs), growing season length decreased with increasing elevationecause of later leaf-out with later leaf-coloring at lower elevation,nd the effect of elevation on leaf-coloring was greater than theffect on leaf-out. For MSF-forbs (i.e., Pa and Pn), growing seasonength was longest at 3200 m due to earlier leaf-out with later leaf-oloring, whereas the shortest growing season at 3400 m resultedrom later leaf-out with earlier leaf-coloring. However, growingeason length was longest at 3400 and 3600 m for Pp and Sa becausef later leaf-out with later/latest leaf-coloring. Although generallyhere was a positive correlation between the duration of the entireeproductive stage and growing season, variation of their ratiosith elevation was species-specific, suggesting that climate change

long the elevation gradient did not affect equally the periods ofeproduction and vegetation for these plants. For example, repro-uctive periods for MSF-forbs were longer at lower elevations,

nd were longest at 3600 m for ESF plants and at 3800 m for Sand Pp (with significant differences between 3200 and 3800 m).ur results show that reproduction was compressed into shorter

ime periods only for Pa and Pn as higher elevations at 3600 and

teorology 189–190 (2014) 220–228 227

3800 m shortened their growing season (Fig. 5). Therefore, similarto previous reports (Sherry et al., 2007), in our study the effects oftemperature on the total reproductive period were species-specific.

High temperatures may speed plant growth through all stagesof development, shortening the duration of stages (Halevy, 1985).However, there are no consistent conclusions about the effect ofwarming on flowering duration, which is species specific (Price andWaser, 1998; Dunne et al., 2003; Post and Forchhammer, 2008;Sherry et al., 2011). We also found that although there were pos-itive correlations between the duration of the entire reproductiveperiod and its component stages (i.e., durations of budding, flower-ing and fruiting) (Table S3), trends in their variation with elevationwere different (Figs. 3 and 4), suggesting that in our study a longerreproductive period did not consistently cause longer durations ofeach reproductive phenological event. Generally, we observed thatwarming (i.e., lower elevation) in winter accelerated bud develop-ment, which shortened bud-set duration for all plants measured,whereas the opposite effects were found for warming during thegrowing season (Table S4). However, warming in winter length-ened flowering duration while warming during the growing seasonshortened flowering duration, except for Pn for which the oppositeeffects were found. Higher temperature in winter only shortenedthe duration of fruit-set for Pa and Pn, but warming during thegrowing season lengthened the duration of fruit-set for Kh, Cs, Paand Pn. The opposite effects of warming were found for Sa.

Water availability may change with warmer temperatures andalso alter reproductive phenology (Giménez-Benavides et al., 2007;Jentsch et al., 2009). Previous studies have demonstrated thatdrought often shortens the duration of flowering in both annualand perennial mesophytes (Dickinson and Dodd, 1976; Steyn et al.,1996; Penuelas et al., 2004; Giménez-Benavides et al., 2007; Alizotiet al., 2010), because flowers are open and lose water through tran-spiration (Galen et al., 1999), but limited soil moisture may slowdevelopment during bud and fruit development stages, lengthen-ing those stages (Galen et al., 1999). However, contrary to previousreports (Sherry et al., 2011), in our study there were inconsistenteffects of soil moisture on phenological events. Higher soil mois-ture prolonged the duration of bud-set for all plants, the durationof flowering for forbs (i.e., Pa and Pn) and the duration of fruit-setfor ESF (i.e., Kh and Cs) and MSF-forbs (i.e., Pa and Pn), whereas itshortened the duration of flowering for graminoids (i.e., Kh, Cs, Paand Pn) and the duration of fruit-set for Pn.

Acknowledgements

This work was supported by funding from the National BasicResearch Program (2013CB956000), Strategic Priority ResearchProgram (B) of the Chinese Academy of Sciences (XDB03030403)and the National Science Foundation of China (41230750 and31272488).

Appendix A. Supplementary data

Supplementary data associated with this article can be found,in the online version, at http://dx.doi.org/10.1016/j.agrformet.2014.01.021.

References

Alizoti, P.G., Kilimis, K., Gallios, P., 2010. Temporal and spatial variation of flower-ing among Pinus nigra Arn. Clones under changing climatic conditions. ForestEcology and Management 259, 786–797.

Archibald, S., Scholes, R.J., 2007. Leaf green-up in a semi-arid African savanna: sep-

arating tree and grass responses to environmental cues. Journal of VegetationScience 18, 583–594.

Arft, A.M., Walker, M.D., Gurevitch, J., Alatalo, J.M., Bret-Harte, M.S., Dale, M., Diemer,M., Gugerli, F., Henry, G.H.R., Jones, M.H., Hollister, R.D., Jónsdóttir, I.S., Laine, K.,Lévesque, E., Marion, G.M., Molau, U., Mølgaard, P., Nordenhäll, U., Raszhivin, V.,

2 st Me

BB

C

C

C

D

D

D

D

D

F

F

F

F

G

G

G

H

HH

I

I

J

K

M

M

M

M

spring phenology on the Tibetan Plateau. Proceedings of the National Academy

28 S. Wang et al. / Agricultural and Fore

Robinson, C.H., Starr, G., Stenström, A., Stenström, M., Totland, Ø., Turner, P.L.,Walker, L.J., Webber, P.J., Welker, J.M., Wookey, P.A., 1999. Responses of tun-dra plants to experimental warming: meta-analysis of the international tundraexperiment. Ecological Monographs 69, 491–511.

arry, R.G., 1981. Mountain Weather and Climate. Methuen, London.ertiller, M.B., Irrisarri, M.P., Ares, J.O., 1990. Phenology of Festuca pallescens in rela-

tion to topography in north-western Patagonia. Journal of Vegetation Science 1,579–584.

huine, I., Morin, X., Bugmann, H., 2010. Warming, photoperiods, and tree pheno-logy. Science 329, 277–278.

olautti, R.I., Eckert, C.G., Barrett, S.C.H., 2010. Evolutionary constraints on adaptiveevolution during range expansion in an invasive plant. Proceedings of the RoyalSociety B: Biology Science 277, 1799–1806.

ook, B.I., Wolkovich, E.M., Parmesan, C., 2012. Divergent responses to spring andwinter warming drive community level flowering trends. Proceedings of theNational Academy of Sciences of the United States of America 109, 9000–9005.

ickinson, C.E., Dodd, J.L., 1976. Phenological patterns in the shortgrass prairie. TheAmerican Midland Naturalist 96, 367–378.

orji, T., Steinr, Ø.T., Mor, S.R., Hopping, K.A., Pan, J., Klein, J.A., 2013. Plant functionaltraits mediate reproductive phenology and success in response to experimentalwarming and snow addition in Tibet. Global Change Biology 19, 459–472.

unne, J.A., Harte, J., Taylor, K.J., 2003. Subalpine meadow flowering phenologyresponses to climate change: integrating experimental and gradient methods.Ecological Monograph 73, 69–86.

unnell, K.L., Travers, S.E., 2011. Shifts in the flowering phenology of the northernGreat Plains: patterns over 100 years. American Journal of Botany 98, 935–945.

urant, J.M., Hjermann, D.Ø., Anker-Nilssen, T., Beaugrand, G., Mysterud, A., Pet-torelli, N., Stenseth, N.C., 2005. Timing and abundance as key mechanismsaffecting trophic interactions in variable environments. Ecology Letters 8,952–958.

itter, A.H., Fitter, R.S.R., Harris, I.T.B., Williamson, M.H., 1995. Relationships betweenfirst flowering date and temperature in the flora of a locality in central England.Functional Ecology 9, 55–60.

itter, A.H., Fitter, R.S.R., 2002. Rapid changes in flowering time in British plants.Science 296, 1689–1691.

orrest, J., Inouye, D.W., Thomson, J.D., 2010. Flowering phenology in subalpinemeadows: does climate variation influence community co-flowering patterns?Ecology 91, 431–440.

ranks, S.J., Sim, S., Weis, A.E., 2007. Rapid evolution of flowering time by an annualplant in response to a climate fluctuation. Proceeding of the National Academyof Sciences of the United States of America 104, 1278–1282.

alen, C., Sherry, R.A., Carroll, A.B., 1999. Are flowers physiological sinks or faucets?Costs and correlates of water use by flowers of Polemonium viscosum. Oecologia118, 461–470.

alloway, L.F., Burgess, K.S., 2012. Artificial selection on flowering time: influence onreproductive phenology across natural light environments. Journal of Ecology100, 852–861.

iménez-Benavides, L., Escudero, A., Iriondo, M.J., 2007. Reproductive limits of alate-flowering high-mountain Mediterranean plant along an elevational climategradient. New Phytologist 173, 367–382.

acker, J., Ladinig, U., Wagner, J., Neuner, G., 2011. Inflorescences of alpine cushionplants freeze autonomously and may survive subzero temperatures by super-cooling. Plant Science 180, 149–156.

alevy, H., 1985. CRC Handbook of Flowering. CRC Press, Boca Raton, FL.ermanand, J.J., Sultan, S.E., 2011. Adaptive transgenerational plasticity in plants:

case studies, mechanisms, and implications for natural populations. Frontiers inPlant Science 2, 1–10.

nouye, D.W., 2008. Effects of climate change on phenology, frost damage, and floralabundance of montane wildflowers. Ecology 89, 353–362.

PCC, 2007. Climate Change 2007: The Physical Science Basis: Summary for Policy-makers. Intergovernmental Panel on Climate Change, Geneva, Switzerland.

entsch, A., Kreyling, J., Boettcher-Treschkow, J., Beierkuhnlein, C., 2009. Beyondgradual warming: extreme weather events alter flower phenology of Europeangrassland and heath species. Global Change Biology 15, 837–849.

örner, C., 1999. Alpine Plant Life: Functional Plant Ecology of High MountainEcosystems. Springer Verlag.

enzel, A., Jakobi, G., Ahas, R., Scheifinger, H., Estrella, N., 2003. Variations of theclimatological growth season (1951–2000) in Germany compared with othercountries. International Journal of Climatology 23, 793–812.

enzel, A., Estrella, N., Testka, A., 2005. Temperature response rates from long-termphenological records. Climate Research 30, 21–28.

enzel, A., Sparks, T.H., Estrella, N.C., Koch, E., Aasa, A., Ahas, R., Alm-kübler, K.,Bissolli, P., Braslavská, O., Chmielewski, F.M., Crepinsek, Z., Curnel, Y., Dahl, A.,Defila, C., Donnelly, A., Ilella, Y., Jatczak, K., Måge, F., Mestre, A., Nordli, Y., Uelas,

J., Ttipirine, P., Ová, V., Scheifinger, H., Striz, M., Susnik, A., Vanvliet, A.J.H., Wiel-golaski, F., Zach, S., Zust, A., 2006. European phenological response to climatechange matches the warming pattern. Global Change Biology 12, 1969–1976.

iller-Rushing, A.J., Primack, R.B., 2008. Global warming and flowering times inThoreau’s Concord: a community perspective. Ecology 89, 332–341.

teorology 189–190 (2014) 220–228

Penuelas, J., Filella, I., Comas, P., 2002. Changed plant and animal life cycles from1952 to 2000 in the Mediterranean region. Global Change Biology 8, 531–544.

Penuelas, J., Filella, I., Zhang, X., Llorens, L., Ogaya, R., Lloret, F., Comas, P., Estiarte,M., Terradas, J., 2004. Complex spatiotemporal phenological shifts as a responseto rainfall changes. New Phytologist 161, 837–846.

Piao, S., Cui, M., Chen, A., Wang, X., Ciais, P., Liu, J., Tang, Y., 2011. Altitude andtemperature dependence of change in the spring vegetation green-up date from1982 to 2006 in the Qinghai-Xizang Plateau. Agricultural and Forest Meteorology151, 1599–1608.

Post, E., Forchhammer, M.C., 2008. Climate change reduces reproductive success ofan Arctic herbivore through trophic mismatch. Philosophical Transactions of theRoyal Society London B: Biology Science 363, 2369–2375.

Post, E.S., Pedersen, C., Wilmers, C.C., Forchhammer, M.C., 2008. Phenologicalsequences reveal aggregate life history response to climatic warming. Ecology89, 363–370.

Price, M.V., Waser, N.M., 1998. Effects of experimental warming on plant reproduc-tive phenology in a subalpine meadow. Ecology 79, 1261–1271.

Schleip, C., Rutishauser, T., Luterbacher, J., Menzel, A., 2008. Time series mod-eling and central European temperature impact assessment of phenologicalrecords over the last 250 years. Journal of Geophysical Research 113, G04026,http://dx.doi.org/10.1029/2007JG000646.

Shen, M., 2011. Spring phenology was not consistently related to winter warmingon the Tibetan Plateau. Proceedings of the National Academy of Sciences of theUnited States of America 108, E91–E92.

Shen, M., Tang, Y., Chen, J., Zhu, X., Zheng, Y., 2011. Influences of temperature andprecipitation before the growing season on spring phenology in grasslands ofthe central and eastern Qinghai-Tibetan Plateau. Agricultural and Forest Mete-orology 151, 1711–1722.

Sherry, R.A., Zhou, X., Gu, S., Arnone III, J.A., Schimel, D.S., Verburg, P.S., Wallace, L.L.,Luo, Y., 2007. Divergence of reproductive phenology under climate warming.Proceeding of the National Academy of Sciences of the United States of America104, 198–202.

Sherry, R.A., Zhou, X., Gu, S., Arnone III, J.A., Johnson, D.W., Schimel, D.S., Verburg,P.S.J., Wallace, L.L., Luo, Y., 2011. Changes in duration of reproductive phases andlagged phenological response to experimental climate warming. Plant Ecologyand Diversity 4, 23–35.

Skelly, D.K., Joseph, L.N., Possingham, H.P., Freidenburg, L.K., Farrugia, T.J., Kinnison,M.T., Hendry, A.P., 2007. Evolutionary responses to climate change. ConservationBiology 21, 1353–1355.

Steyn, H.M., van Rooyen, N., van Rooyen, M.W., Theron, G.K., 1996. The phenologyof Namaqualand ephemeral species. The effect of water stress. Journal of AridEnvironments 33, 49–62.

Visser, M.E., Both, C., 2005. Shifts in phenology due to global climate change: theneed for a yardstick. Proceedings of the Royal Society B: Biology Science 272,2561–2569.

Walker, M.D., Wahren, C.H., Hollister, R.D., Henry, G.H.R., Ahlquist, L.E., Alatalo, J.M.,Bret-Harte, M.S., Calef, M.P., Callaghan, T.V., Carroll, A.B., Epstein, H.E., Jónsdóttir,I.S., Klein, J.A., Magnússon, B., Molau, U., Oberbauer, S.F., Rewa, S.P., Robinson,C.H., Shaver, G.R., Suding, K.N., Thompson, C.C., Tolvanen, A., Totland, Ø., Turner,P.L., Tweedie, C.E., Webber, P.J., Wookey, P.A., 2006. Plant community responsesto experimental warming across the tundra biome. Proceedings of the NationalAcademy of Sciences of the United States of America 103, 1342–1346.

Walther, G.R., Post, E., Convey, P., Menzel, A., Parmesan, C., Beebee, T.J.C., Fromentin,J.M., Hoegh-Guldberg, O., Bairlein, F., 2002. Ecological responses to recent cli-mate change. Nature 416, 389–395.

Wang, G.X., Li, Y.S., Wang, Y.B., Wu, Q.B., 2008. Effects of permafrost thawing on vege-tation and soil carbon pool losses on the Qinghai-Tibet Plateau, China. Geoderma143, 143–152.

Willis, C.G., Ruhfel, B.R., Primack, R.B., Miller-Rushing, A.J., Losos, J.B., Davis,C.C., 2010. Favorable climate change response explains non-native species’success in Thoreau’s woods. PLoS ONE 5, e8878, http://dx.doi.org/10.1371/journal.pone.0008878.

Wolkovich, E.M., Cook, B.I., Allen, J.M., Crimmins, T.M., Betancourt, J.L., Travers, S.E.,Pau, S., Regetz, J., Davies, T.J., Kraft, N.J.B., Ault, T.R., Bolmgren, K., Mazer, S.J.,McCabe, G.J., McGill, B.J., Parmesan, C., Salamin, N., Schwartz, M.D., Cleland, E.E.,et al., 2012. Warming experiments underpredict plant phenological responsesto climate change. Nature 485, 494–497.

Woods, E.C., Hastings, A.P., Turley, N.E., Heard, S.B., Agrawal, A.A., 2012. Adaptivegeographical clines in the growth and defense of a native plant. Ecological Mono-graphs 82, 149–168.

Yang, L.H., Rudolf, V.H.W., 2010. Phenology, ontogeny and the effects of climatechange on the timing of species interactions. Ecology Letters 13, 1–10.

Yu, H., Luedeling, E., Xu, J., 2010. Winter and spring warming result in delayed

of Sciences of the United States of America 107, 22151–22156.Ziello, C., Estrella, N., Kostova, M., Koch, E., Menzel, A., 2009. Influence of altitude on

phenology of selected plant species in the Alpine region (1971–2000). ClimateResearch 39, 227–234.