Download - Active layer thermal regime at different vegetation covers at Lions Rump, King George Island, Maritime Antarctica

Geomorphology xxx (2014) xxx–xxx

GEOMOR-04730; No of Pages 11

Contents lists available at ScienceDirect

Geomorphology

j ourna l homepage: www.e lsev ie r .com/ locate /geomorph

Active layer thermal regime at different vegetation covers at Lions Rump, King GeorgeIsland, Maritime Antarctica

Ivan C.C. Almeida a,b,⁎, Carlos Ernesto G.R. Schaefer b, Raphael B.A. Fernandes b, Thiago T.C. Pereira c,b,Alexandre Nieuwendam d, Antônio Batista Pereira e

a Instituto Federal Norte de Minas Gerais, Campus Januária, Fazenda São Geraldo, CEP 39480-000 Januária, MG, Brazilb INCT-Criosfera, Departamente de Solos, Universidade Federal de Viçosa, Av. PH Rolfs, CEP 36570-000 Viçosa, MG, Brazilc Universidade do Estado de Minas Gerais, Av. Prof. Mário Palmério, 1001 Frutal, MG, Brazild Centro de Estudos Geográficos-IGOT, Universidade de Lisboa, Portugale INCT-APA, Universidade Federal do Pampa, Campus São Gabriel, Av. Antônio Trilha, Centro, CEP 97300-000 São Gabriel, RS, Brazil

⁎ Corresponding author.E-mail addresses: [email protected] (I.C.C. A

(C.E.G.R. Schaefer), [email protected] (R.B.A. Fernandes), tor(T.T.C. Pereira), [email protected] (A. Nieuwendam(A.B. Pereira).

http://dx.doi.org/10.1016/j.geomorph.2014.03.0480169-555X/© 2014 Elsevier B.V. All rights reserved.

Please cite this article as: Almeida, I.C.C., etMaritime Antarctica, Geomorphology (2014

a b s t r a c t
a r t i c l e i n f o
Article history:Received 4 March 2013Received in revised form 24 March 2014Accepted 31 March 2014Available online xxxx

Keywords:Soil thermal regimeClimate changeCryosolPermafrostn-F index

Climate change impacts the biotic and abiotic components of polar ecosystems, affecting the stability of perma-frost, active layer thickness, vegetation, and soil. This paper describes the active layer thermal regimes of two ad-jacent shallow boreholes, under the same soil but with two different vegetations. The study is location in LionsRump, at King George Island, Maritime Antarctic, one of the most sensitive regions to climate change, locatednear the climatic limit of Antarctic permafrost. Both sites are a Turbic Cambic Cryosol formed on andesitic basalt,one under moss vegetation (Andreaea gainii, at 85 m a.s.l.) and another under lichen (Usnea sp., at 86 m a.s.l.),located 10 m apart. Ground temperature at same depths (10, 30 and 80 cm), water content at 80 cm depthand air temperature were recorded hourly between March 2009 and February 2011. The two sites showed sig-nificant differences in mean annual ground temperature for all depths. The lichen site showed a higher soil tem-perature amplitude compared to the moss site, with ground surface (10 cm) showing the highest dailytemperature in January 2011 (7.3 °C) and the lowest daily temperature in August (−16.5 °C). The soil temper-ature at the lichen site closely followed the air temperature trend. Themoss site showed a higherwater content atthe bottommost layer, consistent with the water-saturated, low landscape position. The observed thermal buff-ering effect undermosses is primarily associated with highermoisture onsite, but a longer duration of the snow-pack (not monitored) may also have influenced the results. Active layer thickness was approximately 150 cm atlow-lying moss site, and 120 cm at well-drained lichen site. This allows to classify these soils as Cryosols (WRB)or Gelisols (Soil Taxonomy), with evident turbic features.

© 2014 Elsevier B.V. All rights reserved.

1. Introduction

Amore precise knowledge on the distribution and properties of Ant-arctic permafrost is essential for the cryosphere and life sciences, since itwill represent a major control on ecosystem modification followingclimate-induced changes (Vieira et al., 2010). Permafrost, defined asany subsurface earth materials remaining below 0 °C for more thantwo years, may be profoundly affected by global warming. Permafrost-affected soils normally present an active layer, which is defined as theportion of soil which experiences seasonal thawing and freezing(Brown et al., 2000).

lmeida), [email protected]@yahoo.com.br), [email protected]

al., Active layer thermal regim), http://dx.doi.org/10.1016/j.

The regional climate is a first-order control on permafrost thermalregime, with local microclimate further affecting ground surface tem-peratures, and therefore, the thermal state of the permafrost at meso-and microscales. Permafrost temperatures and distribution depend onclimatic and topographic factors, such as air temperature, solar radia-tion, and snow cover, as related to aspect, slope angle and altitude(Luetschg et al., 2004). At a local scale, ground surface and permafrosttemperatures are affected by the characteristics of the snowpack, thetype and height of vegetation, moisture content in the ground, topogra-phy, the geothermal flux, and the thermal diffusivity of the earthmaterials (Judge, 1973). Hence, permafrost and vegetation are key envi-ronmental components and both are sensitive to climate change(Guglielmin et al., 2008), particularly where permafrost has a discontin-uous distribution (Burgess et al., 2000), as in the case of MaritimeAntarctica (Bockheim et al., 2013) where Lions Rump is located. Mari-time Antarctica has been increasingly recognized as a key region formonitoring climate change (Cannone et al., 2006; Vieira et al., 2010;Bockheim et al., 2013).

e at different vegetation covers at Lions Rump, King George Island,geomorph.2014.03.048

http://dx.doi.org/10.1016/j.geomorph.2014.03.048

mailto:[email protected]







http://www.sciencedirect.com/science/journal/0169555X


2 I.C.C. Almeida et al. / Geomorphology xxx (2014) xxx–xxx

Specific investigations focusing on the relationships between Ant-arctic vegetation and permafrost as a sensitive indicator of climatechange are still lacking (Guglielmin et al., 2005; Cannone et al., 2006;Guglielmin et al., 2008, 2012; Michel et al., 2012).

The aims of this paper are to investigate the influence of vegetationcover on thermal and hygrometric ground regimes, and to describe theactive layer thermal regimeof two shallow boreholes of a Turbic CambicCryosol located at 85m a.s.l., under two adjacent contrasting vegetationtypes: a moss carpet (Andreaea gainii) and a lichen (Usnea sp.)community.

2. Materials and methods

2.1. Regional characteristics

The study sites are located at Lions Rump Peninsula in King GeorgeBay, King George Island, Maritime Antarctica (Fig. 1). Both sites are lo-cated in a soil classified as Turbic Cambic Cryosol (Eutric, Skeletic), ac-cording to the WRB system (IUSS, 2007) (Table 1), which correspondto loamy-skeletal, mixed, subgelic, Typic Haplorthels, according to theSoil Taxonomy (SSS, 2010). Both sites are located at the same altitudeof ≅85 m, have a loamy-sand texture (Table 2) with the only differencerelated to vegetation cover: the moss carpet (mainly formed byA. gainii) site in the low microdepression at a lower position (watersaturated rill), and the lichen community site, dominated by Usneaantarctica and Ochrolechia frigida and located in a slightly higher undu-lated terrain (inter-rill). The vegetation coversmore than 70% of the soilsurface at both sites (Fig. 2).

There are no local meteorological data for Lions Rump. Data on cli-mate acquired at the Brazilian Commandant Ferraz, the nearest stationof Lions Rump, report mean monthly air temperatures varying from−6.4 °C in July to+2.3 °C in February, with mean annual precipitationof 400 mm (Simas et al., 2007). The mean annual air temperature

Fig. 1. Location of the studied sites, in Lions Rump Peninsula, K(Adapted from Michel et al., 2012.)

Please cite this article as: Almeida, I.C.C., et al., Active layer thermal regimMaritime Antarctica, Geomorphology (2014), http://dx.doi.org/10.1016/j.

(MAAT) was −2.6, −1.1 and −2.6 °C for the years 2009, 2010 and2011 respectively. The actual soil water regime is highly influenced bylandscape position, snowpack and the intense snow and permafrostthawing during the summer. The temperature remains above freezingpoint throughout the summer (November to March) allowing plantgrowth, especially mosses, lichens and algae (Schaefer et al., 2004)and the two higher plants Deschampsia antarctica and Colobanthusquitensis.

According to thedata assembled byVieira et al. (2010) during the In-ternational Polar Year (IPY) permafrost temperature monitoring showsthat South Shetland Islands have temperatures slightly below 0 °C atlow elevations, with values decreasing to about −1.8 °C at 270 m a.s.l.and 25m depth. Coastal areas about sea level are essentially permafrostfree (Vieira et al., 2010).

2.2. Methods

In the summer of 2009, two trenches were dug and the active layermonitoring sites were set up in both sites. Themonitoring systems con-sist of soil temperature probes (Campbell L107E thermocouple, accura-cy of ±0.2 °C) arranged in a vertical array at different depths at bothsites (10, 30 and 80 cm) (Table 3); soil moisture probes (CS616 watercontent reflectometer, accuracy of ±2.5%) were placed at the bottom-most layer at each site (80 cm deep in both boreholes) to determinethe volumetric water content; and an air temperature thermistor witha ventilated radiation shield (accuracy of ±0.1 °C) was installed at100 cm above the soil surface to measure temperature (Fig. 2C). Allprobes were connected to a Campbell Scientific CR 1000 dataloggerthat recorded data at hourly intervals from April 2009 to January2011. The main characteristics of the monitored sites are presented inTable 1.

A multivariable regression was performed to describe soil tempera-ture at 80 cm from temperatures of overlying layers and air temperature,

ing George Island, King George Bay, Maritime Antarctica.



Table 1General characteristics of monitored sites.

Site Geographic positiona Altitudem a.s.l.

Vegetation % coverb Soil class WRB/soil taxonomy ALTc

cm

Moss carpet site 04411653109959

85 Andreaea gainii, with sparse tuffs of Deschampsiaantarctica and Colobanthus quitensis

70 Turbic Cambic Cryosol (Eutric. Skeletic)/TypicHaplorthels

147

Lichen site 04411653109959

86 Usnea antarctica and Ochrolechia frigida on deadmosses

80 Turbic Cambic Cryosol (Eutric. Skeletic)/TypicHaplorthels

120

a UTM Zone 21S, WGS 84.b Estimated.c ALT = active layer thickness (estimated).

3I.C.C. Almeida et al. / Geomorphology xxx (2014) xxx–xxx

using Statistica® software (Statsoft, 2009). We calculated the followingparameters: (1) thawing days (days in which all hourly measurementsare positive with at least one reading warmer than +0.5 °C); (2) freez-ing days (days in which all hourly measurements are negative with atleast one reading colder than −0.5 °C); (3) number of isothermal days(days in which all the hourly measurements range only between±0.5 °C); the number of freeze–thaw days (days with both negativeand positive temperatures with at least one value greater than±0.5 °C); (4) thawing degree days (obtained by the cumulative sumof the mean daily temperatures above 0 °C); (5) freezing degree days(obtained by the cumulative sum of the mean daily temperaturesbelow 0 °C); all parameters were obtained according to the available lit-erature (Guglielmin et al., 2008).

To assess the influence of the average snow pack thickness, we esti-mated the N factor index (n-F). The n-F index relates air temperaturewith the ground surface and can be calculated for the thawing degreedays or freezing degree days of the soil (Riseborough, 2003). With thesoil freezing degree days at 10 cm (FDD) and air (FDDa), we calculatedn-F index, according to Eq. (1):

n‐F ¼ FDD=FDDa: ð1Þ

The soil temperature at 10 cmwas selected as the closest to the soilsurface (Nieuwendam, 2009). No directmeasurement of snowpackwasmade.

The active layer thickness was calculated as the 0 °C depth by ex-trapolating the thermal gradient from the two deepest temperaturemeasurements (Guglielmin, 2006). The apparent thermal diffusivitywas estimated according to McGaw et al. (1978) by using Eq. (2):

α ¼ ΔZ2=2Δt� Tijþ1−Tj

i−1� �

= Tji−1−2T1j þ Tj

iþ1� �h i

ð2Þ

where α = apparent thermal diffusivity (m2 s−1), Δt = time incre-ments (s),ΔZ= space increments (m), T= temperature, j= temporalposition and i= depth position. Nelson et al. (1985), Outcalt andHinkel(1989), Hinkel et al. (1990, 2001) andMichel et al. (2012) also used thisparameter to assess the resistance to energy flux in soils. Hourly esti-mates were made for intermediate depths of both profiles, and meanvalues were calculated and plotted for each day (Fig. 3C).

3. Results and discussion

Although both sites are similar in terms of latitude, altitude, soil andlocal climate, the active layer thermal regimewas considerably differentwhen comparing the influence of two covers (Fig. 3). The contrastingbehavior regarding thewater content and soil temperature is attributedto the vegetation cover and the microenvironment at each site. Themoss site is located in a furrow position (hydromorphic)whereas the li-chen site is located on a slightly elevated ridge (well-drained) (Fig. 2).These small differences can cause significant differences in the soilwater/temperature regime.


3.1. Moss carpet site

This site is cover by a 3 cm tall of Andreaeamoss carpet. The highestrecorded soil temperature at this site was 12.1 °C (Table 3), not at thetop soil but at 30 cm (January 20th, 2010 at 2:00 PM), while the highestmean daily air temperature was 11.1 °C recorded on January 3rd,2011 at 4:00 PM. Soil water content ranged from approximately 50%in summer (closer to saturation) to 13% in winter (Fig. 3A). This varia-tion in water content show that summer water saturation above thepermafrost is followed by freezing of active layer, which protects theground from rapid cooling and creates a strong thermal gradient(Romanovsky and Osterkamp, 2000) due to the high latent heat ofwater.

Variations on mean annual temperatures are not observed. Themean annual temperatures are equal along all profile (−0.5 °C). Themaximum and minimum temperatures (hourly records) for the bot-tommost layer (depth of 80 cm) were 7.1 °C and −4.8 °C, respectively(Table 3). This high temperature (7.1 °C) in the 80 cmdepth favors deepactive layer development; in fact, this data stands out as the deepest ac-tive layer depth at King George Island at this altitude, so far recorded.Positive monthly mean temperatures at 80 cm occurred from Januaryuntil April 2010, and started early in the next spring, at November2010. A chill occurs at mid-summer in February, slowing down thesoil-warming trend (Tables 3 and 4).

The permafrost was estimated to start around 150 cm, defined as alayer which remains frozen (below 0 °C) for two consecutive years. Al-though frequent daily freeze and thaw cycles were observed, the activelayer remains predominantly unfrozen until the end of March 2009,when the temperature decreases close to 0.0 °C, coinciding with begin-ning of thermal autumn (Fig. 3C), when the so-called zero curtain re-gime occurs (Hinkel et al., 2001). That keeps the temperature close to0.0 °C until the upper layer is complete frozen, at the end of May. Thezero curtain regime is a result of phase change, and this phenomenonis known to delay the onset of ground freezing during the autumn. Asthe melting front moves downward in the snow pack, meltwater is re-leased. This melt water reaches the cold subsurface of the soil and cre-ates a layer of superimposed ice, where it refreezes and releasesenergy. The length of the zero curtain is dependent on the amount ofsuperimposed ice created, which acts as an energy sink that has to befilled before a warming of the active layer can begin (Hanson andHoelzle, 2004). Upon completion of the phase change, the continual re-moval of heat results in a temperature reduction (Hinkel et al., 2001). Inthe following year, 2010, the active layer remained predominantly un-frozen until April 17th, 2010, when the temperature decreased againclose to 0.0 °C, and a new thermal autumn and zero curtain regimebegan, extending until June 5th, 2010. After this time, temperature de-creased and the upper soil layer started to freeze down to 80 cmdepth at the end of June. During the thermal autumn, the freezingfront did not reach the complete active layer depth. After June 23rd,the soil presented consistent negative temperatures at the 80 cmdepth, indicating complete freezing of the active layer at that time. Fur-ther cooling occurred during the winter season, with soil temperaturesdecreasing to a minimum of −6.5 °C at 10 cm and −4.8 °C at 80 cmdepth.



Table 2Morphological and physical properties of studied profiles from Lions Rump.

Depth(cm)

Structurea Transition TFSA%

CSb FSc Siltd Claye O.M.f Class

dag/kg

Moss carpet siteA 0–8 w l bl/sg Clear wavy 52 57 25 14 4 2.1 Loamy sandAB 8–14 w l bl/sg Gradual wavy 60 56 22 17 5 1.6 Loamy sandBi1 14–26 w m bl/st m gr Gradual irregular 28 62 21 15 2 0.3 Loamy sandBi2 26–48 md st l bl/md m gr Gradual irregular 34 58 20 20 2 0.0 Loamy sandBC 48–80+ md m l bl – 45 56 23 19 2 0.1 Loamy sand

Lichen siteA 0–8 w l bl/sg Clear wavy 52 51 28 18 3 1.9 Loamy sandAB 8–20 w l bl/sg Gradual wavy 50 49 29 18 4 1.0 Loamy sandB1 20–33 w m bl/st m gr Gradual irregular 40 47 29 21 3 0.8 Loamy sandBi2 33–52 md st l bl/md m gr Gradual irregular 52 52 25 21 2 0.1 Loamy sandBC 52–82+ md m l bl – 69 58 22 19 1 0.1 Loamy sand

a Structure classification: Development: w = weak. md = moderate. st = strong. Size: f = fine. m = medium, l = large. Type: ma = massive. gr = granular. bl = subangularblocky. sg = single grain. cr = crumbs.

b Coarse sand (2–0.2 mm). Soil physic was analyzed with texture procedure (EMBRAPA, 1997 modified by Ruiz, 2005).c Fine sand (0.2–0.05 mm). Soil physic was analyzed with texture procedure (EMBRAPA, 1997 modified by Ruiz, 2005).d Silt (0.05–0.002 mm). Soil physic was analyzed with texture procedure (EMBRAPA, 1997 modified by Ruiz, 2005).e Clay (b0.002 mm). Soil physic was analyzed with texture procedure (EMBRAPA, 1997 modified by Ruiz, 2005).f O.M. = organic matter (Walkley–Black method). Soil physic was analyzed with texture procedure (EMBRAPA, 1997 modified by Ruiz, 2005).


The spring thermal regime started at the end of September, with soiltemperature rising rapidly, especially for the topsoil layer, starting at−4.1 °C on August 21st and reaching 0 °C on October 26th. Differentfrom the freezing during autumn, which showed a delay between thetop and bottommost layers, the thawing phenomenon practically oc-curred simultaneously in the whole active layer profile. All year round,the temperature difference between the top and bottommost sensorwas low, reaching a positive maximum of 7.7 °C on December 9th,2009 and negative minimum of −4.3 °C on June 7th, 2010, closely ac-companying the beginning of summer and winter, respectively. LateMarch, June (autumn) and October (middle spring) were periods oflower thermal amplitude, with temperatures in the entire active layerclose to 0 °C.

The water content is highly variable ranging from around 13% in thewinter to 50% in the summer (Fig. 3A). These high contents of soil watercan be explained by the hydromorphic location of the moss site, wherethe meltdown of snow and thawing of active layer continuously supplywater to thismicrodepression. Thewater content followed the zero cur-tain regimes, so that when this soil reaches temperatures above lique-faction the water content increases quickly (Fig. 3A). The same occurswhen the soil starts freezing, i.e., the water content just decreaseswhen soil temperature records are below the freezing point.

The soil thermal regime provides useful data to evaluate the physicalstate of soil moisture and, traditionally, simple approaches have beenadopted providing proxy measures at the phase state of soil moisture(Guglielmin et al., 2008). Among these, the “isothermal days” indicate

Fig. 2. Vegetation cover in Lions Rump: A) Moss carpet si


the energy consumed to change the water phase both during thawingand freezing, when latent heat exchange produces the “zero curtain ef-fect” (Outcalt et al., 1990). In a similar way, the “frozen days” and the“thawed days” indicate when the soil moisture is frozen or in a liquidstate. Moreover, the “freeze–thaw days” describe days in whichfreeze–thaw cycles may occur (Guglielmin et al., 2008).

The thawing days varied from 104 days at 10 cm depth to 123 at thedeeper layers (Table 4), and did not vary much between layers studied.Thawing days are concentrated at early summer and autumn, with atendency of isothermal days to increase with depth, reaching 188 daysat 80 cm. This was expected because permafrost and the higher watersoil content are close, promoting the thermal buffering effect(Table 5). At the depth of 10 cm, two freeze–thaw days occurred;while at the other depths only one freeze–thaw day occurred.

Consistent with Michel et al's. (2012) observations, thawing andfreezing cumulative degree days provide an estimate of the net energyflux in the studied sites. Comparison of the 10 cm thawing degreedays (236 degree days) and freezing degree days (−435 degree days)indicated predominantly freezing conditions (Table 6). At the bottomof the soil, the thawing degree days totaled 191 degree days, whilefreezing degree days totaled−516 degree days, during the monitoringtime (March 2009 to January 2011). However, in one year alone (Dec2009 to Nov 2010) thawing degree days presented 123 degree days,while freezing degree days presented −224 degree days. Thus, forthis soil, freezing conditions are nearly twice as frequent as thawingconditions, during the year (Table 6). Summer's thawing is also affected

te; B) Lichen site; C) Partial view of network sensors.



Table 3Monthly mean temperature (°C) at air and different depths, and water content (WC, %) at80 cm depth from Lions Rump soil under two vegetation covers.

Moss Lichen

Depth (cm) WC Depth (cm) WC

Month Air temp. 10 30 80 % 10 30 80 %

Apr/09 −2.6 −0.1 −0.1 −0.1 29 −1.4 −0.2 0.0 19May/09 −3.7 −0.4 −0.1 −0.1 29 −2.7 −1.5 −0.3 17Jun/09 −8.7 −2.1 −1.2 −0.4 23 −6.5 −4.8 −2.7 13Jul/09 −9.8 −3.5 −2.8 −2.1 16 −8.1 −7.0 −5.5 11Aug/09 −9.4 −5.2 −4.6 −4.1 14 −8.9 −8.3 −7.3 10Sep/09 −4.8 −4.1 −4.0 −3.9 14 −4.5 −4.7 −4.6 11Oct/09 −3.1 −1.4 −1.8 −2.1 15 −1.9 −2.1 −2.2 12Nov/09 −2.6 −1.1 −1.4 −1.7 16 −1.0 −1.8 −2.1 13Dec/09 0.2 0.8 0.4 −0.2 24 1.7 0.0 −0.6 14Jan/10 0.2 1.3 1.0 0.6 44 2.1 0.9 −0.1 15Feb/10 −0.5 0.0 0.0 0.0 43 0.9 0.5 0.0 16Mar/10 −0.4 0.6 0.4 0.2 38 0.4 0.4 0.1 18Apr/10 −3.0 0.1 0.1 0.0 37 −1.4 −0.3 0.0 19May/10 −3.3 −0.4 −0.1 −0.1 37 −2.5 −1.7 −0.7 15Jun/10 −2.6 −1.9 −0.9 −0.3 28 −2.9 −2.3 −1.6 14Jul/10 −4.9 −2.1 −1.7 −1.4 17 −4.8 −4.1 −3.4 13Aug/10 −5.0 −3.6 −3.2 −2.7 15 −5.2 −4.9 −4.3 12Sep/10 −2.9 −3.2 −3.2 −3.0 15 −3.4 −3.5 −3.5 13Oct/10 −0.2 −0.5 −0.9 −1.2 17 −0.1 −0.9 −1.4 14Nov/10 1.2 2.2 1.7 0.9 44 2.0 0.2 −0.4 15Dec/10 −0.2 1.1 0.5 0.2 47 1.9 0.8 0.0 16Jan/11 1.2 2.8 1.8 0.8 39 3.6 2.1 0.8 20Mean −3.0 −0.9 −0.9 −0.9 27 −1.9 −2.0 −1.8 14Min −25.0 −6.5 −5.5 −4.8 13 −17.2 −13.2 −9.9 10Max 11.1 11.2 12.1 7.2 50 13.7 4.6 1.3 24St dev 5.1 2.4 1.9 1.5 13 4.1 2.9 2.2 3


by this buffering effect at the same depth, but with less intensity, possi-bly due to high amounts of ground ice accumulating over the perma-frost table (Michel et al., 2012).

The thermal diffusivity was estimated for the intermediate depth(30 cm) for each hour, reporting as daily mean value (Fig. 3C). Since itwas calculated from the observed temperatures, it includes the thermalimpact of nonconductive heat transfer, and should bemore properly re-ferred to as the apparent thermal diffusivity (ATD) (Hinkel et al., 2001).Mean ATD for a 12 month period was−4.6 × 310−06 m2 s−1. NegativeATD values indicate that nonconductive effects oppose and overcomethe conductive trend (Outcalt and Hinkel, 1989) due to higher watercontent, especially during summertime.

The moss site ATD presented large positive and negative variationsfrom December to April, with drastic fluctuations of soil temperatureand water content and little variations during May and June (Fig. 3C).Due to its relative lower location (micro-depression), it is highly suscep-tible to moisture variations during summer and springtime.

The ATD values are low and positive during late autumn. Cooler au-tumn temperatures result in a near-isothermal soil condition, with littleheat transport and almost no variability of ATD, consistent with Hinkelet al. (2001). The dense moss carpet can function as an insulatingcover during the late autumn, similar to the snow effect during thewin-ter, preventing energy flux, so the mean ATD during winter (21 June to23 September) was positive at 30 cm (1.7 × 10−06 m2 s−1), but ratherlow due to the retarding heat flux, blocking the freezing front. Thisvalue of ATD during the winter was similar to that reported by Michelet al. (2012) at Fildes Peninsula in a soil covered with 50% of mixedvegetation (mosses and lichens). This suggests a variability of thermalbehavior under a given type of vegetation, preventing a generalization.

3.2. Lichen site

The highest soil temperature record on lichen sitewas 13.7 °C in Jan-uary 3rd, 2011 (Table 3), while the highest mean daily air temperaturewas 7.3 °C, at the same date. The smoothing and delaying of the soilthermal regime is greatly enhanced with depth, and the highest


temperature did not exceed 1.3 °C at 80 cm. The water content is littlevariable, ranging from around 10% in the winter to 24% in the summer(Fig. 3B); hence, the lichens-covered soils have much less water thanthe soil under mosses, side-by-side. Mean annual temperatures of−1.1 °C,−1.3 °C and−1.3 °C were obtained for increasing depths, re-spectively at 10 cm, 30 cm, and 80 cm depth between the beginning ofsummer (December 21st) until late spring (December 20th). Positivetemperatures occur at the depth of 80 cm, but the highest record was1.3 °C, whereas the mean monthly temperatures was close to 0.0 °C atthis depth, indicating that it remained close to freezing even duringthe summer. The active layer thickness (ALT) estimated was around120 cm. The soil temperature under lichen cover closely followed thevariability in air temperature, with the upper layers beingmore variable(Fig. 3). In contrast with themoss site that showed a lower temperatureamplitude between the top and bottommost layer, the lichen site had ahigher amplitude, reaching amaximum positive value of 12.3 °C on De-cember 26th, 2010 and negativeminimumof−7.3 °C on July 6th, 2010,corresponding to early summer and mid-winter times, respectively.This high amplitude under lichens is due to high fluctuations at thetop layer, whereas temperature closely followed the air temperaturetrend, whereas the subsurface layers remain buffered because the air-filled porosity (greater stoniness). A greater amount of rock fragmentsis typical of lichen community throughout Antarctica (Longton, 1988).

Negative temperatures at the upper layer started in the last week ofMarch 2009 with the beginning of the thermal autumn, whereas tem-peratures dropped rapidly and remained frozen until middle November(Fig. 3B). In 2010, the upper layer started to cool at the end ofMarch andremained frozen by the end of October. It is important to remark that along zero curtain regime (N40 days) only happened in the bottommostlayer due to the highest water content, and the proximity to permafrost,so that the bottom layer started the cooling process in lateMarch, whilefreezing occurred only at mid-May. In the upper layers, temperatureschanged with air temperature, and no buffering effect (zero curtain re-gime) was verified in these frozen layers when cooling started (Fig. 3B).This is consistent with Guglielmin et al. (2008), who also observed lon-ger zero curtain regime in the bottommost layer compared to the upperones, under a similar Usnea cover at Signy Island.

The process of energy transfer from the atmosphere to the first 10cm of soil had a delay of just 10 h (Fig. 3) and was quite effective atdeeper layers. Temperature reached −17.2 °C at 10 cm, and −9.1 at80 cm with a delay of 30 h.

Thawing started in mid-November 2009 and late October 2010when the upper layer presented positive temperatures. The atmospher-ic heating started at September, with soil temperature in the upperlayers increase exponentially, following the air temperature trend. De-spite the rapid thaw in the upper layer, the temperature at 80 cmdepth remained negative until mid-December, reaching a maximumof 1.3 °C in February 2011, which indicates proximity to the permafrosttable (≅120 cm). These results are similar to those found byMichel et al.(2012) in the nearby Potter Peninsula, who found 0.1 °C at 90 cmdepthat a 70 m a.s.l. soil position, under a similar lichen cover.

The thawing days are concentrated in summertime varying from128 days for 10 cm depth, to 44 days at the 80 cm depth (Table 4).With lower water content at the soil surface the energy that dissipatedin the changing water phase (solid to liquid) is decreased withwarming.

Freezing days occurred year round, except in the summer, beinghigher in the lichen site compared to mosses. The estimate of freezingdays indicates a range of 422 days in the upper layer to 483 days at80 cm depth.

The mean apparent thermal diffusivity (ATD) for the whole studiedperiod was 3.5 × 10−6 m2 s−1 at 30 cm depth. Colder conditions result-ed in lower ATD variability. Mean ATD obtained in winter (from June21st to September 23rd) was positive (6.6 × 10−6 m2 s−1). Contraryto data observed for the moss cover, the ATD values under lichenpresent great positive oscillations from May to June, negative from



Fig. 3. Daily mean soil and air temperatures and water content at A) Moss carpet site; B) Lichen site. C) Apparent thermal diffusivity (ATD) of intermediate depth (30 cm) at both sites.


August to November and late October to early December, and minorvariations betweenMarch and April. Highest ATD values were observedin late autumn, when the soil is dry and still uncovered by snow, so dif-fusivity of the freezing front is high. When winter begins and the soilsurface is already covered by snow cooling is reduced. During the win-ter, soil heat transfer is dominated by conduction, when most of thesoil water has been converted to ice. During springtime, with snowmelting, infiltration of meltwater produces a thermal pulse in the active


layer, and that can significantly hasten soil warming (Hinkel et al., 1997,2001) showing negatives values of ATD. During the summer, withhigher water content, the exchange heat is mainly by conduction.

3.3. Intersite variations

Both sites showed significant differences inmean temperature for alldepths (Table 3). Themoss carpet presented higher water content year-



Table 4Thawing and freezing days (°C) of the studied sites in Lions Rump during a period of 22 months.

Month Moss Lichen Moss Lichen

Thawing days Freezing days

10 cm 30 cm 80 cm 10 cm 30 cm 80 cm 10 cm 30 cm 80 cm 10 cm 30 cm 80 cm

Apr/09 0 0 0 0 0 0 6 5 4 29 15 0May/09 0 0 0 0 0 0 29 19 11 31 31 21Jun/09 0 0 0 0 0 0 30 30 30 30 30 30Jul/09 0 0 0 0 0 0 31 31 31 31 31 31Aug/09 0 0 0 0 0 0 31 31 31 31 31 31Sep/09 0 0 0 0 0 0 30 30 30 30 30 30Oct/09 0 0 0 0 0 0 31 31 31 31 31 31Nov/09 0 0 0 0 0 0 30 30 30 17 30 30Dec/09 10 8 5 18 9 0 8 18 21 0 18 31Jan/10 16 16 16 22 27 0 0 0 0 0 0 19Feb/10 1 1 1 11 13 3 0 0 0 0 0 0Mar/10 13 13 12 9 12 8 0 0 0 11 0 0Apr/10 10 11 9 2 0 0 11 6 6 15 7 2May/10 0 0 0 0 0 0 21 11 9 24 31 31Jun/10 0 0 0 0 0 0 30 25 19 30 30 30Jul/10 0 0 0 0 0 0 31 31 31 31 31 31Aug/10 0 0 0 0 0 0 31 31 31 31 31 31Sep/10 0 0 0 0 0 0 30 30 30 30 30 30Oct/10 2 0 0 1 0 0 26 30 30 17 31 31Nov/10 25 24 21 19 17 0 0 0 2 0 6 30Dec/10 15 14 9 17 16 2 0 0 0 3 0 13Jan/11 31 31 31 29 31 31 0 0 0 0 0 0Sum 123 118 104 128 125 44 406 389 377 422 444 483


round, with greater values during the summer (Fig. 3), due to its lowerlandscape position. The soil temperature under lichen cover closelyfollowed the variability of air temperate.

The soil under lichens exhibits higher number of thawing days at thesurface layers (10 and 30 cm depths) and a lower number of thawingdays in the 80 cm, suggest an insulating behavior by trapped aircompromising the energy transfer to deeper layers when compared tothe soil under mosses (Table 3).

Although the thawing days began in December 2009 in both sites,the moss site showed a warming at the deepest layer in the samemonth,while at the lichen site it only occurred in February 2010, endingin March (Table 3).

Table 5Isothermal and freeze–thaw days (°C) of the studied sites at Lions Rump during a period of 22

Month Moss Lichen

Isothermal days

10 cm 30 cm 80 cm 10 cm 30 cm 80 cm

Apr/09 23 25 26 1 15 29May/09 2 12 19 0 0 10Jun/09 0 0 0 0 0 0Jul/09 0 0 0 0 0 0Aug/09 0 0 0 0 0 0Sep/09 0 0 0 0 0 0Oct/09 0 0 0 0 0 0Nov/09 0 0 0 1 0 0Dec/09 13 4 5 2 4 0Jan/10 15 15 15 4 3 10Feb/10 27 27 28 10 14 25Mar/10 18 20 19 8 19 23Apr/10 9 8 15 2 23 28May/10 10 20 22 4 0 0Jun/10 0 5 11 0 0 0Jul/10 0 0 0 0 0 0Aug/10 0 0 0 0 0 0Sep/10 0 0 0 0 0 0Oct/10 2 1 0 2 0 0Nov/10 5 6 6 0 8 0Dec/10 16 17 22 5 13 16Jan/11 0 0 0 0 0 0Sum 140 160 188 39 99 141


The surface stoniness is similar at both sites (48% of coarsematerialsN 2 mm) (Table 2) and texture did not differ also. The organic mattercontent at the surface was slightly higher at moss site, but otherwise,the two sites were comparable (Table 2).

Due to water lateral and horizontal percolation from snowmelt todeeper layers in the summer, heat exchange between air and soil layeris favored, then more heat flux is translocated to deeper layers princi-pally in the moss site. This process was much reduced at the lichensite due to less moisture, ensuing thermal insulation by air trapped insoil pores. Owing to a water-saturated state at moss site, deeper layersexperiences highest temperature through the summer (Table 3). Inthis case, the small thickness of the moss cover (3 cm) and relatively

months.

Moss Lichen

Freeze–thaw days

10 cm 30 cm 80 cm 10 cm 30 cm 80 cm

0 0 0 0 0 00 0 0 0 0 00 0 0 0 0 00 0 0 0 0 00 0 0 0 0 00 0 0 0 0 00 0 0 0 0 00 0 0 12 0 00 1 0 11 0 00 0 0 5 1 00 0 0 7 0 00 0 0 3 0 01 0 0 11 0 00 0 0 3 0 00 0 0 0 0 00 0 0 0 0 00 0 0 0 0 00 0 0 0 0 01 0 0 19 0 00 0 1 11 0 00 0 0 6 1 00 0 0 2 0 02 1 1 90 2 0



Table 6Thawing and freezing degree days of the studied sites at Lions Rump during a period of 22 months.

Month Moss Lichen Moss Lichen

Thawing degree days (cumulative temp. N 0 °C) Freezing degree days (cumulative temp. b 0 °C)

10 cm 30 cm 80 cm 10 cm 30 cm 80 cm 10 cm 30 cm 80 cm 10 cm 30 cm 80 cm

Apr/09 18 9 7 0 10 17 −12 −21 −23 −30 −20 −13May/09 0 0 0 0 0 0 −31 −31 −31 −31 −31 −31Jun/09 0 0 0 0 0 0 −30 −30 −30 −30 −30 −30Jul/09 0 0 0 0 0 0 −31 −31 −31 −31 −31 −31Aug/09 0 0 0 0 0 0 −31 −31 −31 −31 −31 −31Sep/09 0 0 0 0 0 0 −30 −30 −30 −30 −30 −30Oct/09 0 0 0 0 0 0 −31 −31 −31 −31 −31 −31Nov/09 0 0 0 13 0 0 −30 −30 −30 −17 −30 −30Dec/09 21 12 9 29 12 0 −10 −19 −22 −2 −19 −31Jan/10 31 31 31 30 31 6 0 0 0 −1 0 −25Feb/10 28 28 28 24 28 28 0 0 0 −4 0 0Mar/10 27 31 28 16 28 31 −4 0 −3 −15 −2 0Apr/10 16 0 0 13 21 23 −14 −12 −12 −17 −9 −7May/10 0 0 0 0 0 0 −31 −31 −31 −30 −30 −30Jun/10 0 0 0 0 0 0 −30 −30 −30 −30 −30 −30Jul/10 0 0 0 0 0 0 −31 −31 −31 −31 −31 −31Aug/10 0 0 0 0 0 0 −31 −31 −31 −31 −31 −31Sep/10 0 0 0 0 0 0 −30 −30 −30 −30 −30 −30Oct/10 4 0 0 13 0 0 −27 −31 −31 −18 −31 −31Nov/10 30 30 27 30 20 0 0 0 −3 0 −10 −30Dec/10 30 30 30 26 31 8 −1 −1 0 −5 0 −13Jan/11 31 31 31 31 31 31 0 0 0 0 0 0Sum 236 202 191 225 212 144 −435 −451 −461 −445 −457 −516


low organic matter accumulation, together with other mosses (such asSanionia — see Guglielmin et al., 2008) apparently prevent an efficientbuffering effect, and the water-saturated state became the key factor.

Owing to the late freezing during the winter and late thawing in thespring, themoss site has more isothermal days than the lichen site; thisis attributed to the differential redistribution of melting water in thearea, with a water-saturated landscape under mosses. The isothermaldays are concentrated during the autumn and summer (Table 5) andthe days with temperature close to 0 °C gradually increases in depthat the moss site, it is also higher in all layers compared to the lichensite, which exhibit abruptly increases of isothermal days with depth.In this respect, Guglielmin et al. (2008) also observed a similar trendof isothermal days for mosses at Signy Island, being also concentratedbetween April and October.

In general, freezing–thawing does not occur in the 80 cm layerat either site, but it is more frequent in the lichen ground surface.This indicates a buffering effect under mosses, where little tem-perature fluctuations are phased with the atmosphere. Under li-chens these fluctuations are closely matched by air temperaturechanges.

The freeze–thaw days are completely absent at 80 cm layer in bothsites, indicating that diurnal fluctuations are smoothed. The freeze–thaw cycles at this depth are also strongly reduced because duringspring and fall (the periods in which air temperature more frequentlycrossed the 0 °C threshold) far more heat had been consumed in thewater phase changes causing the temperature fluctuations to besmoothed, as suggested by Guglielmin et al. (2008).

Unlike other studies comparing different vegetation covers(Cannone et al., 2006; Guglielmin et al., 2008) the estimated activelayer thickness was greater at the moss (≅150 cm) than at the lichen(≅120 cm) site, due a lower thermal conductivity under lichens, withdrier soils. Both boreholes did not reach the permafrost table. Similarto that observed by Kane et al. (2001) studying soils in Alaska, in LionsRump the average depth of thaw is greater in the vicinity of water-saturated depressions (moss site) than in inter-rill, well drained areas(lichen site). Under the mosses, the soil remains saturated for thewhole summer season; thus, it attains enhanced thermal conductivitycompared to well-drained areas. This enhanced thermal conductivityeffectively supplies a warm boundary condition for the upper surface


to the mineral soil layer, thereby producing enhanced thaw in these lo-cations, as suggested by Kane et al. (2001).

The Typic Haploturbels studied occur under periglacial conditionsjust above 80 m a.s.l. at Lions Rump, representing Gelisols, with gelicmaterials within 100 cm of the soil surface and permafrost within200 cm of the soil surface; it is inferred that no continuous permafrostwill occur in lower altitudes. Ramos and Vieira (2003) observed perma-frost starting below 230 cm in Livingston Island monitoring a boreholeat 35 m a.s.l. In general, active layer is typically greater than 90 cm forMaritime Antarctica areas as exposed by Vieira et al. (2010) consideringother monitored boreholes in the South Shetlands.

When the soil has enough liquid water available during the thawingand isothermal days (Thorn et al., 2002), chemical weathering processmay occur. Following this reasoning these thermal conditions weremore favorable for chemical weathering in the moss site than in the li-chen site for all soil layers.

Michel et al. (2012) observed that soil texture is amajor influence onsoil energy flux for Antarctic soils, with different textures but similarvegetation cover in Potter and Fildes Peninsula. These authors consid-ered that coarse-textured soils have a considerable influence on energyflux, contributing to permafrost preservation.

In sites with distinct vegetation covers Cannone et al. (2006) report-ed that different vegetations resulted in contrasting buffering effect,snow thickness and longevity. They also remarked that snow cover isthinner and melts earlier on Usnea and Deschampsia sites compared toa pure moss (Sanionia) site.

Temperature oscillations are greater in the lichen site, but after thezero curtain regime the moss site showed temperature changes with alinear curve, while the lichen site revealed changes in harmonic waves(Fig. 3).

The seasonal evolution of the freezing n-factor is presented in Fig. 4.The results indicate differences of freezing duration during the two sea-sons,which started in lateMarch in 2009 andmid-April in 2010.We canalso observe that the n-F values are very different at the two sites, withgreater value for the lichen site.

In 2009, in the beginning of the freezing season, the n-F values fromthe lichen site are rather low (0.2), but increased throughout thewinter,reaching a stable point in late June, with a 0.7 n-F value. From this pointuntil early November the value stayed rather stable. The moss site has a



Fig. 4. N-Factor measured at lichen and moss sites for years 2009 (A) and 2010 (B). The soil temperature at 10 cm was used because it was the closer to the soil surface.


very low n-F value at the beginning of the cold season, but increasedduring thewinter, reaching amaximumvalue of 0.4 in early October, re-maining stable until December.

In 2010 the differences between the two sites were evident. At thelichen site freezing started much later then the Moss site, being alsomuch shorter. The lichen site started freezing in mid-June, showing ann-F value of 0.5, and increasing rather fast, reaching n-F value of 1. Thevalues were variable until September. TheMoss site had a very constantincrease throughout the season, starting againwith ann-F value close to0, and reaching a maximum of 0.5 in November.

These results show two important evidences. First, the snow cover in2009 was apparently thicker than in 2010, because the n-F values werelower in both sites, demonstrating less influence of air temperatures onground temperatures due to thicker snow pack. We may also have aslightly greater snowpack longevity at the concavemoss site, comparedto the lichen site (more exposed to wind ablation), since mosses hadlower n-F values in the whole period.

Although the freezing n-factors vary inter-annually each site has dif-ferent n-factors. The differences can be dominated by variations in snowcharacteristics. The maximum depth and duration of snow cover, aswell as the timing and rate of the snow cover development, have beenshown to greatly influence the ground thermal regime (Ling andZhang, 2003).

Multivariate regression considering the temperature at the low-ermost layer as dependent on air and the other layer's temperaturesgenerated good equation estimates for each site (Table 7). The ad-justed R-squared was higher than 0.91 for both sites, and standard

Table 7Multivariate regression for estimative of deepest layer temperature as dependent of air and othsignificant for all beta values).

Site Equation

Usnea Tsoil 80 cm = 0.010328 × Tair − 0.293544 × Tsoil 10 cm + 1.106034 × Tsoil 3Moss Tsoil 80 cm = 0.007493 × Tair − 0.377956 × Tsoil 10 cm + 1.243124 × Tsoil 3


errors for the beta coefficients were within an acceptable range.The equations help us understand the joint linear effect on the setof independent variables (beta coefficients) in predicting tempera-ture of the bottom (80 cm) of the profile (dependent variable). Asexpected, both sites were more influenced by the closest subse-quent horizon (30 cm), presenting similar beta coefficient; andthe contribution of air and topsoil temperatures (10 cm) wassmall, due to the time lag in temperature changes (Fig. 3A and B).However, this information should be interpreted with care, sincethe higher (top soil) sensor does not reflect the actual soil surfacetemperature.

All beta coefficients were statistically significant. The equation pro-posed for the moss site overestimated temperature peaks in summer,and the predicted values diverged from observed ones by a maximumof 8.0, a minimum of −1.6, and an average of 0.1; during the winterthe temperatures are too close (Fig. 5). This overestimated temperaturein the moss site during the summer was attributed to greater soil mois-ture, so that more heat exchange was needed to change soil tempera-ture because the high water latent heat. In the drier lichen site, valuesestimated from the equations were close to those registered with thesensors, with a uniform behavior throughout the studied period. Resi-dues ranged fromamaximumof 2.80 to aminimumof−3.1with an av-erage of 0.2 (Fig. 5). Observed values at moss are more concentratedbetween −4 and 4 °C and lichen between −10 and 3 °C, a featurethat indicates the buffering effect undermosses, due to the combinationof highermoisture observed and a possible longer duration of the snow-pack, not directly measured in this study.

er layers' temperature (N = 16,778, hourly data, alpha = 5%, p b 0.000000 for both sites

Adjusted R2 Standard error of beta values

0 cm 0.9118 Tair (0.0017); Tsoil 10 cm (0.0036); Tsoil 30 cm (0.0038)0 cm 0.9377 Tair (0.0007); Tsoil 10 cm (0.0042); Tsoil 30 cm (0.0049)



-12

-8

-4

0

4

-12 -8 -4 0 4

Temperature at 80 cm (oC)

Temperature at 80 cm (oC)

-12

-8

-4

0

4

-12 -8 -4 0 4

Est

imat

ed te

mpe

ratu

re a

t 80

cm (o C

)E

stim

ated

tem

pera

ture

at 8

0 cm

(o C)

A

B

Fig 5. Comparison between soil temperaturesmeasured at 80 cmdepthwith that estimat-ed from air and 10 and 30 cm depth temperatures at Lions Rump between April 2009 andJanuary 2011 for A) Lichen site; B) Moss site.


4. Conclusions

1. The active layer thermal regime in the two sites studied at LionsRump, Maritime Antarctica, is typical of periglacial environments,with major variation close to the surface during the summerresulting in frequent freeze and thaw cycles. Most soil temperaturereadings for both sites were close to 0 °C resulting in low freezingand thawing degree days. Both soils have a general low thermal ap-parent diffusivity.

2. There are differences in temperature and moisture regimes betweenmosses and lichen communities.

3. The temperature profile of the soils during the studied period indi-cates that the active layer thickness was approximately 150 cm at


low-lying moss site, and 120 cm at well-drained lichen site. This in-dicates that the permafrost table was within 200 cm from the sur-face, corroborating their classification as Cryosols or Gelisols, withevident turbic features.

4. The observed thermal buffering effect under mosses is primarily as-sociated with higher moisture onsite, but a longer duration of thesnowpack (not monitored) may also have influenced the results.

5. These results highlight the importance of coupledmoisture and tem-perature monitoring of Cryosols for a better understanding of thethermal and hygrometric dynamics of permafrost affected-soilsunder contrasting vegetation cover.

6. Data presented are obtained from less than two years of monitoring,which limits conclusive interpretations in terms of regional warmingor cooling trends, permafrost distribution and effects of climatechange on the studied soils and vegetation.

Acknowledgments

We thank CNPq for the scholarship and financial support. CarlosSchaefer thanks CAPES for granting a sabbatical leave at CambridgeUniversity. We wish to thank Prof. Mauro Guglielmin and Prof. AdrianHarvey for the suggestions and comments that have improved theman-uscript significantly. This is a contribution of the TERRANTAR group(INCT da Criosfera— PROANTAR).

References

Bockheim, J., Vieira, G., Ramos, M., López-Martínez, J., Serrano, E., Guglielmin, M.,Wilhelm, K., Nieuwendam, A., 2013. Climate warming and permafrost dynamics inthe Antarctic Peninsula region. Glob. Planet. Chang. 100, 215–223.

Brown, J., Hinkel, K., Nelson, F.E., 2000. The Circumpolar Active Layer Monitoring (CALM)program: research designs and initial results. Polar Geogr. 24, 165–258.

Burgess, M., Smith, S.L., Brown, J., Romanovsky, V., Hinkel, K., 2000. Global Terrestrial Net-work for Permafrost (GTN-P): permafrost monitoring contributing to global climateobservations. Current Research 2000 — E14. Geological Survey of Canada, pp. 1–8.

Cannone, N., Ellis Evans, C.J., Strachan, R., Guglielmin, M., 2006. Interactions between cli-mate, vegetation and the active layer in soils at two Maritime Antarctic sites. Antarct.Sci. 18, 323–333.

EMPRESA BRASILEIRA DE PESQUISA AGROPECUÁRIA – EMBRAPA, 1997. Manual demétodos de análise de solo, 2.ed. Rio de Janeiro, Centro Nacional de Pesquisa deSolos (212p.).

Guglielmin, M., 2006. Ground surface temperature (GST), active layer and permafrostmonitoring in continental Antarctica. Permafr. Periglac. Process. 17, 133–143.

Guglielmin, M., Ellis Evans, C.J., Cannone, N., 2005. Interactions between climate, vegeta-tion and active layer in Maritime and Continental Antarctica for climate changemon-itoring. Terra Antarct. Rep. 11, 15–27.

Guglielmin, M., Ellis Evans, C.J., Cannone, N., 2008. Active layer thermal regime underdifferent vegetation conditions in permafrost areas. A case study at Signy Island(Maritime Antarctica). Geoderma 144, 73–85.

Guglielmin, M., Wordland, M.R., Cannone, N., 2012. Spatial and temporal variability ofground surface temperature and active layer thickness at the margin of MaritimeAntarctica, Signy Island. Geomorphology 155–156, 20–33.

Hanson, S., Hoelzle, 2004. The thermal regime of the active layer at theMurtèl rock glacierbased on data from 2002. Permafr. Periglac. Process. 15, 273–282.

Hinkel, K.M., Outcalt, S.I., Nelson, F.E., 1990. Temperature variation and apparent thermaldiffusivity in the refreezing active layer, Toolik Lake, Alaska. Permafr. Periglac. Pro-cess. 14, 265–274.

Hinkel, K.M., Outcalt, S.I., Taylor, A.E., 1997. Seasonal patterns of coupled flow in the activelayer at three sites in northwest North America. Can. J. Earth Sci. 34, 667–678.

Hinkel, K.M., Paetzold, F., Nelson, F.E., Bockheim, J.G., 2001. Patterns of soil temperatureand moisture in the active layer and upper permafrost at Barrow, Alaska:1993–1999. Glob. Planet. Chang. 29, 293–309.

IUSS Working Group WRB, 2007. World Reference Base for Soil Resources 2006, FirstUpdate 2007. World Soil Resources Reports. , 103. FAO, Rome (116 pp.).

Judge, A., 1973. The prediction of permafrost thicknesses. Can. Geotech. J. 10, 1–11.Kane, D.L., Hinkel, K.M., Goering, D.J., Hinzman, L.D., Outcalt, S.I., 2001. Non-conductive

heat transfer associated with frozen soils. Glob. Planet. Chang. 29, 275–292.Ling, F., Zhang, T., 2003. Impact of the timing and duration of seasonal snow cover on the

active layer and permafrost in the Alaskan Arctic. Permafr. Periglac. Process. 14,141–150.

Longton, R.E., 1988. The Biology of Polar Bryophytes and Lichens. Cambridge UniversityPress, Cambridge (395 pp.).

Luetschg, M., Stoeckli, V., Lehning, M., Haeberli, W., Ammann, W., 2004. Temperatures intwo boreholes at Fluela Pass, eastern Swiss Alps: the effect of snow redistribution pat-terns in high mountain areas. Permafr. Periglac. Process. 15, 283–297.

McGaw, R.W., Outcalt, S.I., Ng, E., 1978. Thermal Properties of Wet Tundra Soils at BarrowAlaska. Proceedings, Third International Conference on Permafrost. , 1. National Re-search Council of Canada, Ottawa, pp. 47–53.


http://refhub.elsevier.com/S0169-555X(14)00182-2/rf0005




















































Michel, R.F.M., Schaefer, C.E.G.R., Poelking, E.L., Simas, F.N.B., Fernandes Filho, E.I.,Bockheim, J.G., 2012. Active layer temperature in two Cryosols from King GeorgeIsland, Maritime Antarctica. Geomorphology 155–156, 12–19.

Nelson, F.E., Outcalt, S.I., Goodwin, C.W., Hinkel, K.M., 1985. Diurnal thermal regime in apeat-covered palsa, Toolik Lake, Alaska. Arctic 38, 310–315.

Nieuwendam, A.T., 2009. Aspectos do regime térmico do solo da Península de Hurd (IlhaLivingston, Antárctida). Dissertação (Mestrado emGeografia) Universidade de Lisboa,Lisboa, UL (138 pp.).

Outcalt, S.I., Hinkel, K.M., 1989. Night frost modulation of near-surface soil–water ion con-centration and thermal fields. Phys. Geogr. 10, 336–346.

Outcalt, S.I., Nelson, F.E., Hinkel, K.M., 1990. The zero-curtain effect: heat and mass trans-fer across an isothermal region in freezing soil. Water Resour. Res. 26, 1509–1516.

Ramos, M., Vieira, G., 2003. Active layer and permafrost monitoring in Livingston Island,Antarctic. First results from 2000 to 2001. In: Phillips, M., Springman, S.M., Arenson,L. (Eds.), Proceedings of the 8th International Conference on Permafrost. Balkema,Lisse, Zurich, pp. 929–933.

Riseborough, D., 2003. Thawing and freezing indices in the active layer. First results from2000 to 2001. In: Phillips, M., Springman, S.M., Arenson, L. (Eds.), Proceedings of the8th International Conference on Permafrost. Balkema, Lisse, Zurich, pp. 953–958.

Romanovsky, V.E., Osterkamp, T.E., 2000. Effects of unfrozen water on heat and masstransport processes in the active layer and permafrost. Permafr. Periglac. Process.11, 219–239.

Ruiz, H.A., 2005. Incremento da exatidão da análise granulométrica do solo por meio dacoleta da suspensão (silte + argila). Revista Brasileira de Ciência do Solo 29, 297–300.


Schaefer, C.E.G.R., Simas, F.N.B., Albuquerque-Filho, M.R., Michel, R.F.M., Viana, J.H.M.,Tatur, A., 2004. Fosfatização: Processo de formação de solos na Baía do Almirantadoe implicações ambientais. In: Schaefer, C.E.G.R., Francelino, M.R., Simas, F.N.B., Albu-querque Filho, M.R. (Eds.), Ecossistemas Costeiros e Monitoramento Ambiental daAntártica Marítima, Baía do Almirantado, Ilha Rei George. NEPUT — Departamentode Solos, Viçosa, Brasil, pp. 47–59.

Simas, F.N.B., Schaefer, C.E.G.R., Melo, V.F., Albuquerque-Filho, M.R., Michel, R.F.M.,Pereira, V.V., Gomes, M.R.M., Costa, L.M., 2007. Ornithogenic cryosols from MaritimeAntarctica: phosphatization as a soil forming process. Geoderma 138, 191–203.

Soil Survey Staff (SSS), 2010. Keys to Soil Taxonomy. USDA, Natural Resources Conserva-tion Service, Eleventh edition. National Soil Survey Center, Lincoln, NE (338 pp.).

Statsoft, 2009. Statistica for Windows. Release 9. Statsoft, Inc., Tulsa, OK.Thorn, C.E., Darmody, R.G., Allen, C.E., Dixon, J.C., 2002. Near-surface ground temperature

regime variability in selected microenvironments, Kärkevagge, Swedish Lapland.Geogr. Ann. 84A, 289–300.

Vieira, G., Bockheim, J., Guglielmin, M., Balks, M., Abramov, A.A., Boelhouwers, J., Cannone,N., Ganzert, L., Gilichinsky, D.A., Goryachkin, S., López-Martínez, J., Meiklejohn, I.,Raffi, R., Ramos, M., Schaefer, C.E.G.R., Serrano, E., Simas, F.N.B., Sletten, R., Wagner,D., 2010. Thermal state of permafrost and active-layer monitoring in the Antarctic:advances during the International Polar Year 2007–2009. Permafr. Periglac. Process.21, 182–197.








































