J Paleolimnol

DOI 10.1007/s10933-013-9703-0

ORIGINAL PAPER

Changes in sub-fossil chironomid assemblages

in two Northern Patagonian lake systems associated
with the occurrence of historical fires
Alberto Araneda • Patricia Jana • Carolina Ortega • Fernando Torrejón •
Sébastien Bertrand • Patricia Vargas • Nathalie Fagel • Denisse Alvarez •
Alejandra Stehr • Roberto Urrutia

Received: 14 October 2011 / Accepted: 18 November 2012

Ó Springer Science+Business Media Dordrecht 2013

Abstract Patagonia is commonly seen as an excep- patterns or increasing nutrient inputs. In the studied
tionally pristine area because of its wildlife and lakes the periods with higher prevalence of fires were
practically unpolluted waters. However, during the identified by charcoal analysis, while organic matter
twentieth century the burning of natural forests was and magnetic susceptibility allowed the confirmation
one of the most important human activities in Northern of pre-fire and post-fire periods. The chironomid
Chilean Patagonia. Some estimations indicate that composition was evaluated through a PCA and an
three million hectares were burned during the first analysis of similarity (ANOSIM) to test the significance
three decades of the century. Hence the objective of among periods while a Detrended Correspondence
this study was to evaluate the impacts of the historical Analysis was applied to the chironomid assemblage
fires in Lake Burgos (458420 S) and Lake Thompson downcore to assess compositional structure and taxa
(458380 S) in Chilean Patagonia. The impact was turnover. In Lake Burgos the ANOSIM test indicated
measured by evaluating chironomid assemblage since significant differences between the pre-fire and fire
they are sensitive enough to be used as an indicator of periods (p \ 0.05), while in Lake Thompson differ-
aquatic ecosystem health. Fires have a direct and ences were not significant. However, in Lake Thompson
drastic effect on a lake watershed but also indirectly the PCA clearly separated the pre-fire from the fire
affect a lake ecosystem, changing sedimentation period but not the fire from the post-fire periods. In both
lakes chironomid composition changed in relation to the

A. Araneda (&)  P. Jana  C. Ortega  N. Fagel

F. Torrejón  P. Vargas  D. Alvarez  R. Urrutia Clays and Paleoclimate Research Unit, Department
Group of Paleolimnological Studies (GEP), Aquatic of Geology, University of Liege, Liège, Belgium
Systems Research Unit, Environmental Sciences Center
EULA-Chile, University of Concepcion, A. Stehr
Concepción, Chile Environmental Engineering Research Unit,
e-mail: aaraneda.ec@gmail.com Environmental Sciences Center EULA-Chile, University
of Concepcion, Concepción, Chile
S. Bertrand
Renard Centre of Marine Geology, University of Ghent,
Krijgslaan 281, S8, 9000 Ghent, Belgium

S. Bertrand
Marine Chemistry and Geochemistry, Woods Hole
Oceanographic Institution, 360 Woods Hole Road, Woods
Hole, MA 02543, USA

123
 J Paleolimnol

period of higher prevalence of fires, which in turn chironomids, have rarely been investigated (Francis
implies catchment changes, pollution, and other anthro- 2001; Tremblay et al. 2010).
pogenic impacts. Particularly a marked change in One of the most frequently used approaches to track
mesotrophic/eutrophic taxa was detected, reflecting an the fire history of a watershed is the analysis of
increase in nutrient input due to deforestation. Our charcoal stored in lake sediments (Long et al. 1998;
findings point out that the lacustrine ecosystems are still Markgraft et al. 2007; McWethy et al. 2010; Whitlock
affected by the impact of fires and the subsequent 2001). During a forest fire the ignition of woody
increase in nutrient supply that occurred almost 50 years biomass produces charcoal particles that are trans-
ago. No sign of reverting to pre-disturbance conditions ported by wind to lacustrine ecosystems, preserved in
was observed, which makes these lakes highly sensitive the bottom sediments (Whitlock et al. 2008). It is
to current human-induced impacts. possible to distinguish two kinds of charcoal: micro-
charcoal (\125-lm) and macrocharcoal ([125-lm;
Keywords Fires impacts  Lake sediments  Markgraf et al. 2007; Whitlock and Larsen 2001).
Chironomids  Charcoal  Southern Chile Unlike microcharcoal, macrocharcoal represents fires
that occurred near the lake or in the basin because they
cannot be dispersed more than 30 km (Clark 1988;
Introduction Thevenon et al. 2010; Whitlock and Larsen 2001). It is
therefore possible to identify past local fire events by
Patagonia has traditionally been described as a very counting macrocharcoal and looking for peaks of its
pristine area with important reserves of wildlife concentration in lacustrine sediment sequences (Long
(Brooks et al. 2006) and a large reservoir of freshwater et al. 1998; Higuera et al. 2010; Philibert et al. 2003).
resources (Rignot et al. 2003). Nevertheless, northern Chironomidae (Insecta: Diptera) comprises one of
Patagonia is currently threatened by increasing human the most abundant, diverse and representative aquatic
activities related to the development of the region insects being their larval stages an important link in
(Vince 2010). However human activities in Patagonia the ecology of aquatic environments (Paggi 2001). In
are not new, and its ecosystems have been impacted by all the larval stages chironomid develop a chitinous
human activities over the past 80 years (Martinic head capsule that is generally very well preserved in
2005). Settlers started to colonize the Aysén region at lake sediments being very good indicators of past
the beginning of the twentieth century causing large environmental changes (Walker et al. 1997). In
fires that destroyed three million hectares between southern South America chironomids have been
1930 and 1950 (Martinic 2005). Hence the forest was successfully used as paleoclimate indicators, espe-
progressively replaced by artificial prairies and after cially for detecting cold events in the Taitao peninsula
1975 by some exotic plantations (Quintanilla 2005). (NW Patagonia; Massaferro and Brooks 2002) and
Fires are probably one of the most important and also in recognizing the Huelmo Mascardi Cold
abrupt disturbances that can occur in a watershed. The Reversal (HMCR; Massaferro et al. 2009). In addition,
direct effect of fire is obviously the loss of native they have been used to track the impact of tephra
vegetation, which increases runoff. Runoff enhances deposition on lacustrine ecosystems (Araneda et al.
soil erosion, which promotes the input of sediment to 2007).
aquatic ecosystems (Markgraf et al. 2007; Whitlock In regards to anthropogenic impacts, chironomids
2001). Among the indirect effects are the changes in have been used to infer trophic evolution of lakes
water quality, mainly associated with a higher input of (Brooks and Birks 2001; Heiri and Lotter 2003), past
nutrients, which can lead to lake eutrophication. levels of oxygen in lake waters (Quinlan and Smol
Eutrophication can rapidly affect the aquatic biota 2001) and as quantitative indicators of salinity
and the functioning of the aquatic system itself (Eggermont et al. 2006). In New Zealand McWethy
(McWethy et al. 2010; Philibert et al. 2003). Some et al. (2010) detected a noticeable change in chiron-
studies have reported changes in diatom assemblages omids associated with fires caused by the first
after the occurrence of fires, due to changes in pH and European settlers. In Argentinean Patagonia Massa-
nutrients (Philibert et al. 2003). Besides, changes ferro et al. (2005) found that the main fluctuations in
induced by fires on other aquatic organisms, like chironomid assemblages in Lake Morenito were in

123
J Paleolimnol

response to human-related environmental distur- from 3,000 mm in the coastal zone (Puerto Aysén) to
bances of the first settlers established in the area, due 611 mm in the eastern side of the Andes (Balmaceda
to induced changes in the trophic state of the lake. airport; Romero 1985). The annual average precipitation
But what is the rationale of studying the response of of the study area is approximately 1,350 mm (Coyhaique
biological proxies to environmental changes in Pata- city), which mainly falls as snow.
gonia? As mentioned above, Patagonian ecosystems The first study site, Lake Burgos (45°420 3500 S;
are undergoing increased pressure for development 72°120 5300 W), is located 25 km to the southwest of
purposes. Estimating the response of these ecosystems Coyhaique (sector ‘‘Seis Lagunas’’) at an elevation of
to current environmental changes requires a good 379 m a.s.l. It has a surface area of 0.23 km2 and a
knowledge of the impact of historical fires. In addition, maximum depth of 34 m (Fig. 1). The vegetation
investigating the way in which fires can affect the around the lake is composed of patches of Nothofagus
chemistry of aquatic ecosystems and the global carbon pumilio (Poepp. & Endl.) Krasser (Gajardo 1994).
dynamics is important since the frequency and inten- Most of the watershed is currently used for agriculture,
sity of fires is expected to increase as a consequence of supporting farms and cattle activities. The second
global warming (Philibert et al. 2003). lake, Lake Thompson (45°380 2600 S; 71°470 0700 W), is
Therefore, the goals of this study are (1) to identify located 20 km to the southeast of Coyhaique, at an
the composition of the chironomid assemblages in the elevation of 750 m a.s.l. (Fig. 1) and is right on the
sedimentary records of two Patagonian lakes; (2) to Chile-Argentina border. It has a surface area of
compare these results with the post-disturbance period 1.18 km2 and a maximum depth of 15 m (Fig. 1).
to determine if there is an effective response of The vegetation in this area is mainly composed of
chironomid assemblages to forces such as fire, and (3) Nothofagus antartica (G. Forst.) Oerst and Berberis
to assess the resilience capacity of lake ecosystems to buxifolia Lam., which are characteristic of the forest-
such impacts. steppe ecotone (Gajardo 1994). Dead trees originating
from the historical fires of the early 1900s are still
Site description visible today in the Lake Thompson watershed. The
watershed is currently used for minor cattle activity
The study area is located in Northern Patagonia in the and important reforestation activity using exotic
Aysén region of Chile. Climate in the area corresponds to species (Pinus ponderosa, P. Lawson & C. Lawson).
the supratemperature belt (Amigo and Ramı́rez 1998), Both lakes catchments are mainly composed of
characterized by humid conditions that present a steep basaltic and andesitic lavas and breccias although
precipitation gradient from the wet western to the dry andesitic and rhyolitic pyroclastic rocks can also occur
eastern side of the Andes. Annual precipitation ranges (Sernageomin 2003).

Fig. 1 a Map of the study

area showing the location of
lakes Burgos and
Thompson. b Lake
Thompson bathymetry.
c Lake Burgos bathymetry

123
 J Paleolimnol

Settlement and fire history Chronology

Settlement of Northern Chilean Patagonia in the The chronology of cores Thompson 10 and Burgos 06
Aysén river basin started at the beginning of the were based on correlations to previously published age
twentieth century with the arrival of livestock com- models obtained on parallel cores Thompson 08E and
panies (Martinic 2005). Pomar (1923) published a map Burgos 07 (Bertrand et al. 2012). Correlations between
of the colonization in Northern Patagonia, which the cores were based on magnetic susceptibility and
shows that the main settlements occurred in areas near organic content (LOI) profiles. The age models of
Puerto Aysén, Coyhaique and Balmaceda. In the sediment cores Thompson 08E and Burgos 07 were
1930s the settlement process regained vigor following based on radiocarbon results only due to low activity
government pressure in the law ‘‘Ley de Colonización of other radionuclides (210Pb and 137Cs). A detailed
de Aysén, N8 485500 , dictated in 1930 (Bizama et al. description of the age models is in Bertrand et al.
2011). (2012).
Since the territory was mainly covered by forest,
the settlers had to deforest it in order to transform it Charcoal
into prairies for cattle grazing (Quintanilla 2005). The
easiest and cheapest way to deforest was by burning. In Lake Burgos charcoal was analyzed from samples
The method consisted starting fires during the spring- from core Burgos 06, which was later correlated with
summer season, after which the settlers accumulated the dated core Burgos 07 (Fig. 2a). In Lake Thomp-
and dried biomass like branches and trees to initiate son, charcoal data were obtained from core Thomp-
the fires in the next season. Settlers controlled the son-09, which was correlated to the dated core
frequency of fires, if the previous fire had not provided Thompson-08 (Fig. 2b). In both cases magnetic sus-
the desired results, the settlers would light a new fire in ceptibility and LOI were used for cores correlation.
the same area the following year. With the increase in Sediment for charcoal analysis were sub-sampled at
the number of settlers, fires became widespread and 1 cm intervals taking two or three ml of wet sediment
continued until the 1950s (Martinic 2005). Bizama which were disaggregated in a hot solution of KOH
et al. (2011) estimated that these fires provoked a loss (10 %) for 20 min at 70 °C (Lynch et al. 2003) and then
of 23 % of the original vegetation. sieved through a 125-lm mesh (Long et al. 1998;
Whitlock 2001). Whitlock et al. (2008), indicated that
the most representative interval of charcoal size is
Methods between 0.125 and 0.250 mm, therefore, we decided to
work with[0.125-mm fraction and not measure the size
Sampling and sedimentological analysis of the particles individually. This fraction was deposited
in a Bogorov counting tray and particles were identified
Sediment cores were retrieved from the deepest part under a stereomicroscope at 329 magnification. Mac-
of the two selected lakes using an Uwitec gravity rocharcoal concentration was estimated by dividing the
corer. The core length reached 77 cm in Lake Burgos number of particles by the volume of the sediment
(core Burgos 06) and 172 cm in Lake Thompson (core sample. In order to determine the charcoal accumulation
Thompson 10). Once in the laboratory sediment cores rates (CHAR, in particles cm-2 year-1), the concentra-
were split lengthwise for non-invasive analysis, tion (particles cm-3) was divided by the deposition rate
including X-ray radiography, magnetic susceptibility (cm year -1; Long et al. 1998).
and visual description. Later the cores were sub-
sampled at 1-cm intervals for analysis of organic Chironomids
content, chironomids and charcoal counting. The
organic content of the sediment was estimated by Chironomids were analysed in cores Burgos 06 and
loss on ignition (LOI) following the method described Thompson-10 correlated with the dated core Burgos
by Boyle (2002). Magnetic susceptibility was mea- 07 using magnetic susceptibility and LOI (Fig. 2).
sured at 1-cm intervals using a Bartington MS2E Then four ml of wet sediment were deflocculated in
sensor. 10 % KOH for 15 min at 708 C and passed through a

123
J Paleolimnol

Fig. 2 Correlation between cores dated and cores analyzed through magnetic susceptibility and LOI. a Lake Burgos b Lake Thompson

123
 J Paleolimnol

90-lm sieve. The[90 lm fraction was later sorted in interest (Fig. 2). In Lake Burgos magnetic susceptibility
a Bogorov counting tray. The head capsules were shows important fluctuations from 77 to 25 cm, ranging
picked out with entomological tweezers, mounted in between 70 and 150 SI units (Fig. 2). A peak reaching
Hydromatrix, and identified using a Zeiss microscope 5,000 SI units at 43 cm most likely represents a tephra
(25, 40, or 1009), following to Wiederholm (1983), layer. From 25 cm to the surface, magnetic susceptibility
Paggi (2001) and Brooks et al. (2007). The relative is more stable than downcore, although an increase is
abundance of each taxon was presented as a percent- observed above 8 cm. Organic matter shows a similar
age of the total abundance in each centimeter using pattern but it is reversed (Fig. 3a); it reaches a maximum
TILIA and TILIA GRAPH and C2 v1.5 (Grimm 1987; of 46 % between 27 and 12 cm. A similar trend was
Juggins 2007). observed in Lake Thompson (Fig. 3b). From 160 to
51 cm magnetic susceptibility values are high (max 4,500
SI units) and organic matter content is low (*10 %),
Statistical analysis
while above 51 cm magnetic susceptibility decreases and
organic matter increases.
To distinguish different associations along the profile,
a Stratigraphically Constrained Sum-of-Squares clus-
ter analysis (CONISS) was applied to the percentage Charcoal
values of the chironomid assemblages using TILIA and
TILIA GRAPH (Grimm 1987) and C2 v1.5 (Juggins Figure 5 represent the Charcoal Accumulation Rate
2007) programs. The significance of changes among (CHAR) in Lake Burgos. Along the sediment record,
the periods was done with an analysis of similarity the charcoal particles occur only in the zone B-III
(ANOSIM, Clarke and Warwick 2001) implemented in (1980–2007). Charcoal particles in Lake Thompson
the package Vegan in (Oksanen et al. 2005). To occur in zones T-II (1880–1990) and T-III (1990–2006).
distinguish possible groupings among the samples The zone T-II has the highest abundance of CHAR in the
corresponding to the pre-fires, fires and post-fires part of the sediment record that corresponds to the fire
period, a Principal Components Analysis (PCA), was period, peaking approximately between 1940 and 1960;
performed using CANOCO v. 4.5 (ter Braak and in agreement with the fires provoked by settlers. CHAR
Smilauer 2002). A Detrended Correspondence Anal- in zone T-III presents homogenous values of charcoal
ysis (DCA) was also applied to the chironomid particles that would represent the post-fire period in the
assemblages in order to assess the compositional most recent part of the sediment record.
structure and taxa turnover throughout the profile
(Birks 1998; Langdon et al. 2004), using the package
Chironomid assemblages
Vegan which was also used to estimate the Shannon
diversity index of the assemblage (Oksanen et al.
Lake Burgos
2005). The Shannon index was selected because it is a
very widely used index for comparing diversity bet-
In order to stabilize the variance and reduce the
ween various habitats (Clarke and Warwick 2001).
‘‘noise’’ of data, further analyses were based on
square-root transformed percentages. To avoid the
influence of rare taxa, only those taxa with abundances
Results higher than 2 % and present in at least two samples
were considered (Brooks 2000; Millet et al. 2007). A
Sediment properties and dates total of 2,096 head capsules were found along the
70 cm sediment profile of Lake Burgos, correspond-
The sediment record of Lake Burgos spans the last ing to 31 taxa. The Orthocladiinae subfamily repre-
157 years in the first 77 cm, giving a temporal resolution sented 41.9 % of the total abundance, followed by
of 2.0 years cm-1. In Lake Thompson the sediment Chironominae with 40.1 % (25.4 % Tribe Chirono-
record of the 172 cm extends back to the last 1,570 mini, 14.7 % Tribe Tanytarsini), Tanypodinae with
cal. years. The first 51 cm span the last 177 years, giving 14.4 % and Podonominae with 3.6 %. At the taxa
a temporal resolution of 3.5 years cm-1 for our period of level, the most abundant were Parakiefferiella (18 %),

123
J Paleolimnol

Fig. 3 Sedimentological
properties of sediment cores
lakes Burgos 06 (a) and
Thompson 10 (b), including
magnetic susceptibility,
organic matter content
(%LOI) chironomids head
concentrations (head ml-1)
and total chironomid head
capsules

Phaenopsectra (12.8 %), Ablabesmyia (7.7 %) and As this very general classification of Tanytarsini most
Limnophyes (6.6 %). Head capsule concentration was likely comprises more than two types, a research effort
always above 5 ml-1 in the entire profile, though the is ongoing to obtain a more clear identification of such
concentration showed a noticeable increase from morphotypes (Massaferro, personal communication).
28 cm to the upper part of the core, averaging 28 per CONISS of chironomid assemblages show three
ml-1 (Fig. 3a). different zones spanning the last 157 years (Fig. 5).
Two different groups of Tanytarsini (types A and
B) were distinguished in the chironomid assemblages, Zone B-I (70–50 cm; 1870–1910 AD)
mainly based on differences in the shape of antennal
pedestal (Fig. 4). An antennal pedestal with a low and Head capsule concentrations in this zone were always
rounded spur characterizes Type A, whereas Type B above 14 capsules ml-1; however, in some samples of
has an antennal pedestal with a longer rounded spur. just 5 ml were more than 80 heads. The most abundant

123
 J Paleolimnol

Fig. 4 Ventral view of some chironomid larvae. a Tanytarsina type A, showing an antennal pedestal with no spur. b Tanytarsina type B
with a prominent spur. c Chironomus plumosus-type and d Chironomus anthracinus-type

Fig. 5 Relative frequencies of chironomid assemblages of Lake Burgos. Charcoal profile, diversity index H0 , Semiterrestrial/Littoral
chironomids, DCA scores and CONISS are also shown

taxon is Parakiefferiella, which presents a maximum fairly constant, averaging a value of 2.4; however,
of 25 % and is fairly constant along the entire zone. there are some fluctuations at the beginning of the zone
Phaenopsectra is the second most abundant taxon. It (Fig. 5).
presents a peak in abundance (20.8 %) in the lower
part of the zone, and it shows a slight decline towards Zone B-II (50–14 cm; 1910–1980 AD)
the upper part of the zone. Other important taxa in this
zone are Tanytarsina B and Ablabesmyia, which are This zone represents the maximum abundance of head
omnipresent but never exceed 16 % of the total capsules in the entire sedimentary profile, reaching a
abundance. Apsectrotanypus and Riethia are always peak of 38 heads ml-1 at 28 cm. In this zone
below 14 % of the abundance. Shannon diversity is Parakiefferiella is the most abundant taxon, peaking

123
J Paleolimnol

(36 %) at 24 cm and decreasing rapidly towards the B-III. Tanytarsina A reaches 30.6 % in zone B-III,
top of the zone. Phaenopsectra shows its maximum while Ablabesmyia increases from 4.2 to 17 %
amount at 50 cm (28.8 %), after which it decreases (Fig. 5). Apsectrotanypus shows a similar trend,
until the end of the zone (12.1 %). Limnophyes shows increasing from 1.6 to 12.8 %. Simultaneously, Cric-
a fluctuating but globally increasing trend, with levels otopus increases from 0 to 8.3 % and Polypedilum
close to 4 % in the lower part of the zone and reaching increases from 2.1 to 6.4 %. In zone B-III, Parapsec-
17.8 % at the end of the zone. Other taxa like Riethia trocladius shows a relatively stable presence, averag-
and Ablabesmyia show some peaks at 46 and 10 cm, ing 7 %, which is very similar to the values obtained
with 14.9 and 17.7 %, respectively. Tanytarsina B for zone B-I. Despite the changes described in each
decreases and in the end of the zone its abundance zone, diversity does not show drastic changes through-
reaches 0 %. Diversity fluctuates noticeably in com- out the profile.
parison with the previous zone, reaching a minimum
of 2.04 at 24 cm. Lake Thompson

Zone B-III (14–0 cm; 1980–2005 AD) A total of 1,885 head capsules from nineteen distinct
taxa were retrieved from the 158 cm long sediment
This section represents the most recent part of the sequence (between 159 and 172 cm only a few head
record, which also contains the maximum amount of capsules were found). The Chironominae subfamily
charcoal particles (Fig. 5). In the lower part of this zone represents 88.7 % (77 % Tribe Chironomini, 11.7 %
the number of head capsules is high and decreases Tribe Tanytarsini) of the total quantity, followed by
toward the upper part of the zone. The lowest quantities Tanypodinae (7.2 %), Orthocladiinae (4.0 %) and
were recorded at 12 cm (7 heads ml-1). There is also a Podonominae (0.1 %). At the taxa level, the most
noticeable decrease in taxa that were abundant in the abundant were Chironomus plumosus (48 %), Riethia
previous zones, such as Parakiefferiella, which shows (14.2 %), and Tanytarsina B (9.2 %).
low but stable levels. Limnophyes also present low The concentration of head capsules shows notice-
numbers, decreasing from 12.8 % in the lower part of able changes along the sequence. In the deeper part of
the zone to 2.3 % in its upper part. Phaenopsectra and the core, the concentration is very low. At times it is
Eukiefferiella are practically absent in this zone except below 5 ml-1 (Fig. 6), and there are less than 20 heads
for the sample at 6 cm, where it presents 1.6 % of the total per sample. The concentrations above 53 cm
total fossil abundance. In the case of Stictochironomus, show a marked increase, reaching a maximum of 20
it is absolutely absent in this zone. heads ml-1 at 8 cm. Checked with species accumula-
Some taxa that were relatively rare or even absent tion curves the amount of 20 head capsules just catches
in zones B-I and B-II show a noteworthy increase in a small proportion of the diversity, complicating

Fig. 6 Relative frequencies of chironomid assemblages of Lake Thompson. Charcoal profile, diversity index H0 , Mesotrophic/
Eutrophic chironomids, DCA scores and CONISS are also shown

123
 J Paleolimnol

inferences about climate or earlier human intervention. 32.8 % at 1 cm). Parachironomus and Alotanypus
For this reason, and because the main objective of the also exhibit small increases in comparison to the
study was to assess recent environmental changes, only previous zone. Nonetheless, levels of C. plumosus
the sequence from 53 to 0 cm (with a higher amount of remain high, although they show a decreasing trend
heads) was considered for further analysis. In this towards the upper part of the zone, where C. anthrac-
segment the CONISS analysis distinguished three inus also becomes important. The diversity index
different zones, which represent approximately the last shows a general increasing trend toward the upper part
178 years. The zones can be described as follows of the zone.
(Fig. 6).
Response of chironomids to fire periods
Zone T-I (51–40 cm; 1830–1880 AD)
According to the charcoal profile, the fires period in
Head capsule concentrations present low values in the Lake Burgos is found between 14 and 0 cm
lower part of this zone, reaching a maximum at (1980–2005 AD; Fig. 5). Only the pre-fire and fire
46–48 cm. In this zone, the most abundant taxon is periods are recorded in Lake Burgos, while the post-
Tanytarsina B, with a level of 41 % at 49 cm. fire period is absent. It is possible to detect a noticeable
Polypedilum, Apedilum, Parakiefferiella, and Tany- change in the assemblage composition between the
tarsina A all show percentages lower than Tanytarsina pre-fire and the fire periods best represented by the
B, even though in this zone they show their overall DCA scores (Fig. 5) and by the PCA biplot of Lake
maximum quantity. Diversity in zone T-I is the highest Burgos (Fig. 7). The ANOSIM test delivered a R value
of the entire profile, reaching a value of 2.21 at 43 cm.

Zone T-II (40–10 cm, 1880–1980 AD)

Head capsule amounts remain very stable in this zone and

are always above 50 per sample, averaging 8 heads ml-1.
This zone also marks an increase in taxa that had low
levels in the previous zones, such as C. plumosus and
Riethia, which reach a level of 87.3 % at 30 cm. Taxa
like Polypedilum, Tanytarsina A and B and Parakieffe-
riella show an opposite trend in this zone, decreasing.
Other taxa that are present in this zone, although low in
concentration, are Ablabesmyia, Limnophyes and Cric-
otopus. Chironomid diversity is highly fluctuating in this
zone, alternating between low (0.99) and high values
(1.96). In contrast to the previous zone, zone T-II is
characterized by a decrease in diversity (Fig. 6).

Zone T-III (10–0 cm, 1980–2007 AD)

This zone represents the most recent part of the

sediment record and according to the amount of
charcoal and the core chronology; it would correspond
to the ‘‘post-fire’’ period. Head capsule concentration
is relatively stable, though it decreases slightly
towards the end of the zone. There is a noticeable
Fig. 7 Biplot of the PCA analysis of chironomids records from
change in this zone, which is marked by the appear-
lakes Burgos and Thompson. The open squares represent the
ance and increase of taxa that were previously absent, pre-fire periods, the filled dots represent the fire periods and the
such as Chironomus anthracinus (maximum of open dots represent the post-fire periods

123
J Paleolimnol

of 0.88 and a p value of 0.001 at 95 % confidence these taxa for mesotrophic/eutrophic conditions
intervals (a = 0.05), indicating that there is a statis- (Brooks et al. 2007; Millet et al. 2007; Tremblay
tically significant difference in the chironomid assem- et al. 2010). This suggests that the eutrophication of
blages between both periods. Lake Thompson started at the beginning of the
In Lake Thompson charcoal data indicate that the twentieth century and is still ongoing. Current water
fire period is represented at 30–15 cm, corresponding quality data of this lake reinforce such a statement
to the years 1930–1980 AD, distinguishing the pre- because the amount of total nitrogen (0.49 mg l-1)
fire, fire and post-fire periods. Chironomid assem- and total phosphorous (0.04 mg l-1, Araneda unpub-
blages also show important changes along the sedi- lished data) are indicative of eutrophic conditions.
ment sequence (Fig. 6), but unlike Lake Burgos, the Despite changes in trophic-related taxa, it is also
DCA scores show the most important change in the possible to distinguish a different proportion of littoral/
lower part of the core and not in the period semi-terrestial taxa between the two lakes. Lake Burgos
corresponding to the fire period. The PCA biplot on shows a relatively stable abundance of Limnophyes and
chironomids and samples (Fig. 7) clearly separates the Parakiefferiella, both are typical for littoral zones, and
fauna of the pre-fire period from those representing the the latter is common in cold waters (Walker et al. 1992,
fire period. However, it fails in distinguishing the post- Brooks et al. 2007). However, in Lake Thompson both
fire period from the fire period. The ANOSIM test taxa have very low abundances, occurring mainly in the
gave a R value of 0.1, which is not significant at 95 % lower part of the core and decreasing toward the recent
confidence intervals (p value = 0.07). past. Such differences could be explained by two factors
that can act in parallel: lake morphometry and inter-
specific competition. Lake Burgos has a proportionally
Discussion higher amount of littoral zones than Lake Thompson,
which could lead to a higher abundance of the association
Chironomid composition Limnophyes/Parakiefferiella and also semiterrestrial
taxa (Fig. 5). With respect to competition, if the historical
According to Brooks (2000), Massaferro et al. (2005) fires were able to trigger a rapid eutrophication process in
and others therein, there have been very few studies of Lake Thompson, a rapid ecological succession probably
Patagonian chironomids. This complicates any eco- took place as well, represented by the dominance of the
logical interpretation due to the scarce knowledge of association C. plumosus/C. anthracinus.
the ecology of chironomid taxa of southern South
America and the lack of modern calibration sets Response of chironomids to fire impacts
(Verschuren and Eggermonta 2006). However, as
(Massaferro and Brooks 2002) indicates, previous There are few studies that evaluate the response of
researchers (Brundin 1966; Saether 1975) showed that chironomids to fires that occurred in lacustrine water-
the ecological typology of chironomid genera has sheds. To our knowledge, Francis (2001), McWethy
worldwide validity. This makes it possible to construct et al. 2010), and Tremblay et al. (2010) are the only
tentative inferences of southern South America, based studies that have addressed this issue until now.
on the ecology of Holartic fauna, which is well known. At our study sites the forest fires that occurred in the
Chironomid species composition and the diversity lake watersheds are clearly represented by the char-
index of Lake Burgos and Lake Thompson show some coal profiles. Lake Burgos only recorded the pre-fire
differences. Lake Burgos recorded a higher number of and fire periods, while in Lake Thompson the pre-fire,
taxa (25) than Lake Thompson (16) implying a higher fire and post-fire periods are clearly represented
value of the Shannon diversity index (2.44 vs. 1.59). (Figs. 6, 7). Lake Burgos has a proportionally bigger
However, one of the most noticeable differences in the watershed than Lake Thompson (Ad/Ao; ratio of
chironomid fauna between the two lakes is the high the surface area of the drainage basin versus surface
proportions of C. plumosus and C. anthracinus in Lake area of the lake; 21 for Lake Burgos versus 13 for
Thompson above 40 cm and the complete absence of Lake Thompson), but after the fires its watershed was
both taxa in Lake Burgos. Ecological data from the rapidly converted to pastures and used for livestock.
Northern Hemisphere indicate a clear preference of This would explain why the sediment record does not

123
 J Paleolimnol

give evidence of the post-fire period. But despite some contrast, the most important change in the DCA1 score
differences in morphometric parameters, the climatic of Lake Thompson occurs in the deepest part of the
settings in both lakes are rather similar (rainfall, core and not in the period with higher fires occur-
winds). Therefore the main factor explaining the rences. This means that another factor was able to
difference in charcoal profile between the lakes is the generate a more sensitive response of the chironomid
timing of the settlement process. Our results suggest assemblages of Lake Thompson. Below 51 cm the
that large forest fires started approximately in 1920 AD number of chironomids decreases rapidly and also the
in the region of Lake Thompson, while they appear to sediment properties change drastically (Fig. 3),
have started much later in the region of Lake Burgos decreasing organic content and increasing magnetic
(Bizama et al. 2011; Figs. 5, 6). It is generally believed susceptibility. This suggests that this sediment could be
that the first settlers started to colonize from the sea composed of glacial clays (Fig. 3), probably associated
(Martinic 2005). Some evidence, however, suggests with the manifestation of a Little Ice Age Type Event,
that a parallel ‘‘informal colonization’’ started from the which generated a noticeable impact on chironomid
Argentinean side (Martinic 2005). Due to the easier composition. Nevertheless, chironomids do not seem to
environmental conditions of the eastern side of North- reflect colder temperatures below 51 cm (absence of
ern Chilean Patagonia—the open and windy pampa— taxa adapted to low temperature). This issue needs
the colonization from Argentina could have been faster further investigation, possibly by increasing the count-
and more efficient. Since Lake Thompson is located ing of head capsules for that period using larger samples.
close to the border of Argentina, it is expected that the Another explanation of why chironomids are not
settlers impacted its watershed earlier and more reflecting a cold period could be due to the low
intensively than the watershed of Lake Burgos. taxonomic resolution of this study, where some taxa
Another factor that could have be responsible for are grouped under the same type but probably some
the high prevalence of forest fires caused by the genera could have different temperature optima. Such a
settlers were the dry periods that characterized the problem hopefully will be solved as the Southern South
beginning of the twentieth century in Patagonia (Holz America taxonomy improves—in fact some effort is
and Veblen 2011). However, the chironomid assem- underway (Massaferro unpublished data), and by
blage in both lakes does not seem to reflect a possible developing specific or local training set for temperature.
increase in temperature. In Lake Thompson Abla- However, increasing taxonomic resolution also implies
besmyia, Apsectrotanypus, Polypedilum and Para- some risks. Walker (2001) and Brooks et al. (2007)
chironomus can be interpreted as warm adapted taxa argue that there is an optimum with respect to taxonomic
(Massaferro et al. 2009), but only Parachironomus resolution, if this is too low it can cause a loss of
increase slightly in the upper part of the core and is information, but an increase in the apparent taxonomic
also defined as an eutrophic taxon, then is not possible resolution can also produce misidentification.
to clearly determine if its changes are due to an Regarding differences among periods, the PCA
increase in nutrients or an increase in temperature. In ordination for Lake Thompson clearly shows a
Lake Burgos Polypedilum, Apsectrotanypus, Abla- difference between the fauna associated to the pre-
besmyia, Labrundinia, Parachironomus, and Cricoto- fire period and the fauna associated to the fire period;
pus can be indicative of warm conditions (Massaferro however, no difference is observed between the fire
et al. 2009). Nonetheless, only Cricotopus shows a and post-fires period (Fig. 7). This suggests that
slight increase in the upper part of the core, but despite charcoal profile indicates that the fire period
Williams et al. (2012) suggested that some genus of has ended; the lake and its biota are still experiencing
Cricotopus do not respond to temperature. Then it the repercussions of the provoked changes. The
cannot be inferred an influence of current global ANOSIM test indicated that changes in chironomids
warming on the chironomid assemblage of both lakes. of Lake Thompson seem to be less well expressed
Specific changes in chironomid assemblages are than those of Lake Burgos. There is however, a clear
clearly observed in the DCA1 scores (Figs. 5, 6). In change towards mesotrophic/eutrophic taxa (C. plumosus,
Lake Burgos the largest change in the DCA1 score C. anthracinus) in the post-fire period, reinforcing the
occurs in the phase with the highest concentrations of idea that there was a higher nutrient input in this lake
charcoal, during the main fire period (Fig. 5). In after the fires.

123
J Paleolimnol

The fact that chironomid assemblage is not return- Scrimgeour et al. (2001), also indicate that benthic
ing to its pre-fires state in Lake Thompson could also be communities impacted by fires returns to pre-distur-
due to the nutrient enrichment after the fires, because bance levels within 15–20 years.
probably was of such magnitude that it generates a total McWethy et al. (2010), who relate chironomid
dominance of eutrophic taxa and thus prevented any responses and forest dynamics indicate that the native
change in the assemblage composition. Nevertheless, forests of New Zealand had little or no history of fire
Brooks et al. (2007) indicate that C. plumosus-type is when European settlers arrived. They observed that
characteristic of strongly eutrophic lakes, while C. after the most severe period of forest fires the recovery
anthracinus-type is more common in moderately of the native forest was slow (centuries). This suggests
eutrophic lakes. Therefore, the behavior of these two the high vulnerability of this forest (formed also by
taxa in the more recent part of the core could be Nothofagus spp.) to repeated fire. These observations
indicative of an incipient nutrient reduction in the lake. are similar to our results from Patagonia, where forests
Our results of Lake Burgos show a difference in formed by N. pumilio and N. betuloides [Mirbel]
chironomid assemblages associated to the pre-fire and Oersted, K., present low natural restoration from the
fire periods, differing from the results of Tremblay fires caused almost 50 years ago (Quintanilla 2005).
et al. (2010), who suggest that fires do not have any Although some studies show that there was a natural
effects on chironomid assemblages. We assume that background of fires before the arrival of the twentieth
this difference is due to the magnitude and extent of century settlers in Patagonia, we hypothesize that the
the anthropogenic fires. In the study area of Tremblay intensity of these early fires, which were also related to
et al. (2010) the fires had an extent of 950–2,000 ha. human occupation (Tehuelches Indians), was of much
By contrast, three million ha of North Patagonian lower magnitude and frequency than those of the
forest were burned during the settlement, with some twentieth century (Huber and Markgraf 2003; Markg-
watersheds being entirely burned using repeated forest raf et al. 2007; Veblen et al. 1999), whose impacts
fire ignition techniques (settler descendants, personal surpassed the resilience capacity of the affected lakes.
communication; Quintanilla 2005). Technical reports Our results of Lake Burgos, therefore, indicate that
(IREN 1979) of the Aysén area indicate that the chironomids respond sensitively to fire impacts. How-
watershed of Lake Burgos was completely burned ever, it is not possible to determine the resilience of the
after the big fires but was rapidly covered by pasture lake system to such changes since the charcoal profile
for livestock. In contrast, one-third of the Lake shows that fires were halted only recently, and chiron-
Thompson watershed was burned, and even today is omid communities do not reflect a change associated to
possible to observe dead trees, suggesting that the this period. The situation in Lake Thompson is a little
latter lake underwent a continuous nutrient input that different: anthropogenic fires were heavily reduced
leads to a noticeable current eutrophication process. during the last 20–30 years, but the chironomid fauna
Francis (2001) also found important changes in the has not yet returned to pre-fire conditions. In fact, it
chironomid assemblages of Douglas Lake, Michigan, continues to show the eutrophication process without
that coincides with the arrival of European settlers. The any sign of decline. Thus, it is possible to state that fire
most distinctive change was an increase of eutrophic impact considerably surpassed the resilience capacity of
taxa like Chironomus, Dicrotendipes, Glyptotendipes, Lake Thompson since approximately 50 years after the
and Polypedilum, with a concomitant decrease in the impact it still undergoing impact inertia. This means that
abundance of Sergentia, which is described as a these Patagonian lakes are highly sensitive to anthro-
mesotrophic taxon that tolerates moderate oxygen pogenic activities of the last decades, and that even
deficiencies. These authors attribute such changes to small changes—compared to the historical impacts—
heavy logging and fires that impacted littoral commu- are limiting the return to pre-disturbance conditions.
nities along with siltation and nutrient enrichment.
With respect to the resilience of lakes, and espe-
cially of chironomids, Heiri and Lotter (2003) indicate Conclusions
that chironomid assemblages display a considerable
resilience to human-induced changes such as cattle Subfossil chironomid records of two lakes in North-
activity, quickly returning to the pre-impact state. western Patagonia allow an evaluation of the impacts

123
 J Paleolimnol

of the anthropogenic forest fires that occurred at Araneda A, Torrejon F, Aguayo M, Torres L, Cruces F, Cis-
beginning the twentieth century. In both sites chiron- ternas M, Urrutia R (2007) Historical records of San Rafael
glacier advances (North Patagonian Icefield): another clue
omids seem to respond sensitively to the impacts to ‘Little Ice Age’ timing in southern Chile? Holocene
generated by the forest fires, although significant 17:987–998
differences between pre-fire and fire periods are only Bertrand S, Araneda A, Vargas P, Jana P, Fagel N, Urrutia R (2012)
recorded in Lake Burgos. As indicated by other Using the N/C ratio to correct bulk radiocarbon ages from lake
sediments: insights from Chilean Patagonia, Quat Geochro-
studies, the main effect of fires on lacustrine ecosys- nol. http://dx.doi.org/10.1016/j.quageo.2012.06.003
tems is to increase the nutrient input. This change is Birks H (1998) Numerical tools in palaeolimnology—progress,
clearly registered in the chironomid assemblage of potentialities, and problems. J Paleolimnol 20:307–332
Lake Thompson, which is dominated in its most recent Bizama G, Torrejón F, Aguayo M, Muñoz MD, Echeverrı́a C,
Urrutia R (2011) Pérdida y fragmentación del bosque na-
part by C. plumosus/C. anthracinus, characteristic of tivo en la cuenca del rı́o Aysén (Patagonia-Chile) durante el
mesotrophic/eutrophic conditions. Present-day levels siglo XX. Revista de Geografı́a del Norte Grande 49:
of nutrients in the lake water confirmed such eutro- 125–138
phication state. In Lake Thompson, the DCA scores Boyle JF (2002) Inorganic geochemical methods in palaeolim-
nology. In: Last W, Smol J (eds) Tracking environmental
indicate that non fire-related factors affected the change using lake sediments. Springer, New York, pp 83–141
chironomid assemblages at the beginning of the Brooks SJ (2000) Lateglacial fossil midge (Insecta: Diptera:
record. According to our preliminary data, this could Chironomidae) stratigraphies from the Swiss Alps. Palae-
be due to the input of glacial sediment to the lake, ogeogr Palaeoclimatol Palaeoecol 159:261–279
Brooks SJ, Birks HJB (2001) Chironomid-inferred temperatures
probably during the Little Ice Age chronozone or from Lateglacial and Holocene sites in north-west Europe:
another Neoglacial period, but to recognize clearly the progress and problems. Quat Sci Rev 20:1723–1741
impact of such event on chironomids assemblages Brooks TM, Mittermeier RA, da Fonseca GAB, Gerlach J,
requires to develop local transfer function for temper- Hoffmann M, Lamoreux JF, Mittermeier CG, Pilgrim JD,
Rodrigues ASL (2006) Global biodiversity conservation
ature. On the other hand, chironomids also suggest that priorities science. Science 313:58–61
the impact of fires in the recent part of the core was of Brooks S, Langdon P, Heiri O (2007) The identification and use
such magnitude that the resilience capacity of the lake of Palaearctic Chironomidae larvae in palaeoecology. QRA
ecosystems was surpassed. Our data suggest that Technical Guide No. 10. Quaternary Research Association,
London, p 276
Patagonian watershed have not yet recovered from the Brundin L (1966) Transantarctic relationships and their signif-
historical fires, indicating that those ecosystems had icance, as evidenced by chironomid midges, with a
little resilience to the severe burns that occurred at monograph of the subfamilies Podonominae and Aph-
beginning of the twentieth century. Finally, it is roteniinae and the austral Heptagyiae. Kungl Svenska
Vetens Handlingar 11:1–472
possible to conclude that these lakes and their Clark J (1988) Particle motion and the theory of charcoal
watersheds are still under the effect of the fire impact, analysis: source area, transport, deposition, and sampling.
which makes them highly sensitive to the present-day Quat Res 30:67–80
human occupation, and that this continued anthrop- Clarke KR, Warwick RM (2001) Change in marine communi-
ties: an approach to statistical analysis and interpretation,
ization precludes a return to pre-fire conditions. 1st edn. Plymouth Marine Laboratory, Plymouth, UK; 2nd
edn. PRIMER-E, Plymouth, UK
Acknowledgments Funding for this research is from the Eggermont H, Heiri O, Verschuren D (2006) Subfossil Chiro-
Fondecyt project N 11080158 and partially from Fondecyt nomidae (Insecta: Diptera) as quantitative indicators for
project 1120765. Also a bilateral cooperation project between past salinity variation in African lakes. Quat Sci Rev
Chile and the Wallonie community of Belgium is acknowledged. 25:1966–1994
Authors are especially grateful to the cooperation of Corporación Francis D (2001) A record of hypolimnetic oxygen conditions in
Nacional Forestal (CONAF) and Carabineros de Chile. We wish a temperate multi-depression lake from chemical evidence
to thank the associate editor and three anonymous reviewers for and chironomid remains. J Paleolimnol 25:351–365
their constructive comments and suggestions. Gajardo R (1994) La Vegetación Natural de Chile Clasificación
y Distribución Geográfica. Editorial Universitaria, Santi-
ago de Chile, p 165
References Grimm E (1987) A Fortran 77 program for stratigraphically
constrained cluster analysis by the method of incremental
Amigo J, Ramı́rez C (1998) A bioclimatic classification of sum of squares. Comput Geosci 13:13–35
Chile: woodland communities in the temperate zone. Plant Heiri O, Lotter AF (2003) 9000 years of chironomid assemblage
Ecol 136:9–26 dynamics in an Alpine lake: long-term trends, sensitivity to

123
J Paleolimnol

disturbance, and resilience of the fauna. J Paleolimnol Paggi A (2001) Diptera: Chironomidae. In: Fernández HR,
30:273–289 Domı́nguez E (eds) Guı́a para la determinación de los
Higuera PE, Whitlock C, Gage JA (2010) Linking tree-ring and artrópodos bentónicos Sudamericanos. Editorial Universi-
sediment-charcoal records to reconstruct fire occurrence taria de Tucumán, Tucumán, Argentina, pp 167–194
and area burned in subalpine forests of Yellowstone Philibert A, Prairie YT, Carcaillet C (2003) 1200 years of fire
National Park, USA. Holocene 21:327–341 impact on biogeochemistry as inferred from high resolution
Holz A, Veblen TT (2011) The amplifying effects of humans on diatom analysis in a kettle lake from the Picea mariana-
fire regimes in temperate rainforests in western Patagonia. moss domain (Quebec, Canada). J Paleolimnol 30:167–181
Palaeogeogr Palaeoclimatol Palaeoecol 3(311):82–92 Pomar J (1923) La concesión del Aysén y el valle Simpson.
Huber UM, Markgraf V (2003) European impact on fire regimes Imprenta Cervantes, Santiago de Chile, p 137
and vegetation dynamics at the steppe-forest ecotone of Quinlan R, Smol JP (2001) Chironomid-based inference models for
southern Patagonia. Holocene 13:567–579 estimating end-of-summer hypolimnetic oxygen from south-
IREN (1979) Perspectivas de desarrollo de los recursos de la central Ontario shield lakes. Freshw Biol 46:1529–1551
Región Aysén. Corporación de Fomento de la Producción, Quintanilla V (2005) Estado de recuperación del bosque nativo
Santiago, Chile, p 507 en una cuenca nordpatagónica de Chile, perturbada por
Juggins S (2007) C2 version 1.5: software for ecological and grandes fuegos acaecidos 50 años atrás (448–458 S). Re-
palaeoecological data analysis and visualisation. Newcas- vista de Geografı́a del Norte Grande 34:73–92
tle upon Tyne: department of geography, University of Rignot E, Rivera A, Casassa G (2003) Contribution of the Pat-
Newcastle. Newcastle upon Tyne, pp 73 agonia icefields of South America to sea level rise. Science
Langdon PG, Barber KE, Lomas-Clarke SH (2004) Recon- 302:434–437
structing climate and environmental change in northern Romero H (1985) Geografı́a de los Climas. In: López E (ed)
England through chironomid and pollen analyses: evidence Geografı́a de Chile. Instituto Geográfico Militar, Santiago
from Talkin Tarn, Cumbria. J Paleolimnol 32:197–213 de Chile, p 243
Long CJ, Whitlock C, Bartlein PJ, Millspaugh SH (1998) A Saether OA (1975) Neartic chironomids as indicators of lake
9000-year fire history from the Oregon Coast Range, based on typology. Verhandlungen der Internationalen Vereinigung für
a high-resolution charcoal study. Can J For Res 28:774–787 Theoretische und Angewandte Limnologie 19:3127–3133
Lynch JA, Clark JS, Bigelow NH, Edwards ME, Finney BP Scrimgeour GJ, TW M, PC A, G C (2001) Benthic macroin-
(2003) Geographic and temporal variations in fire history in vertebrate biomass and wildfires: evidence for enrichment
boreal ecosystems of Alaska. J Geophys Res 108:8-1–8-17 of boreal subarctic lakes. Freshw Biol 46:367–378
Markgraf V, Whitlock C, Haberle S (2007) Vegetation and fire Sernageomin (2003) Mapa Geológico de Chile: versión digital.
history during the last 18000 cal year B.P. in Southern Servicio Nacional de Geologı́a y Minerı́a, Publicación
Patagonia: mallı́n Pollux, Coyhaique, Province Aysén Geológica Digital, 4 (CD-ROM, versión1.0, 2003), Santi-
(45°410 3000 S, 71°500 3000 W, 640 m elevation). Palaeogeogr ago de Chile, pp 23
Palaeocl 254:492–507 ter Braak CJF, Šmilauer P (2002) CANOCO reference manual
Martinic M (2005) De la Trpananda alAysén. Pehuén Editores, version 4.5. Microcomputer Power, Ithaca, NY
Santiago de Chile, p 539 Thevenon F, Williamson D, Bard E, Anselmetti FS, Beaufort L,
Massaferro J, Brooks SJ (2002) Response of chironomids to late Cachier H (2010) Combining charcoal and elemental black
quaternary environmental change in the Taitao Peninsula, carbon analysis in sedimentary archives: implications for
southern Chile. Quat Res 17:101–111 past fire regimes, the pyrogenic carbon cycle, and the
Massaferro J, Brooks SJ, Haberle SG (2005) The dynamics of human–climate interactions. Glob Planet Chang 72:381–389
chironomid assemblages and vegetation during the late Tremblay V, Larocque-Tobler I, Sirois P (2010) Historical
quaternary at Laguna Facil, Chonos Archipelago, southern variability of subfossil chironomids (Diptera: Chironomi-
Chile. Quat Sci Rev 24:2510–2522 dae) in three lakes impacted by natural and anthropogenic
Massaferro JI, Moreno PI, Denton GH, Vandergoes M, Dief- disturbances. J Paleolimnol 44:483–495
fenbacher-Krall A (2009) Chironomid and pollen evidence Veblen TT, Kitzberger T, Villalba R, Donnegan J (1999) Fire
for climate fluctuations during the last glacial termination history in Northern Patagonia: the roles of humans and
in NW Patagonia. Quat Sci Rev 28:517–525 climatic variation. Ecol Monogr 69:47–67
McWethy DB, Whitlock C, Wilmshurst JM, McGlone MS, Verschuren D, Eggermonta H (2006) Quaternary paleoecology
Fromont M, Li X, Dieffenbacher-Krall A, Hobbs WO, Fritz of aquatic Diptera in tropical and Southern hemisphere
SC, Cook ER (2010) Rapid landscape transformation in regions, with special reference to the Chironomidae. Quat
South Island, New Zealand, following initial Polynesian Sci Rev 25:1926–1947
settlement. Proc Natl Acad Sci USA 107 Vince G (2010) Dams for Patagonia. Science 329:382–385
Millet L, Vannière B, Verneaux V, Magny M, Disnar J, Lag- Walker IR (2001) Midges: Chironomidae and related diptera. In:
goun-Défarge F, Walter-Simonnet A, Bossuet G, Ortu E, Smol JP, Birks HJB, Last WM (eds) Tracking environ-
de Beaulieu J (2007) Response of littoral chironomid mental change using lake sediments. Volume 4: Zoological
communities matter to late glacial lake-level, vegetation indicators. Kluwer, Dordrecht, p 217
changes at Lago dell’Accesa (Tuscany, Italy). J Paleolim- Walker IR, Oliver DR, Dillon ME (1992) The larva and habitat
nol 38:525–539 of Parakiefferiella Nigra Brundin (Diptera: Chironomi-
Oksanen JR, Kindt P, Legendre, O’Hara RB (2005) Vegan: dae). Neth J Aquat Ecol 24:527–531
community ecology package version 1.15-3. (http://cc.oulu. Walker IR, Levesque AJ, Cwynar LC, Lotter AF (1997) An
fi/*jarioksa/) expanded surface-water paleotemperature inference model

123
 J Paleolimnol

for use with fossil midges from eastern Canada. J Paleo- change from northern Yellowstone National Park based on
limnol 18:165–178 a comparison of multiple proxy data. Quat Int 188:126–138
Whitlock C (2001) Variations in Holocene fire frequency: a Wiederholm T (1983) Chironomidae of the holartic region.
view from the western United States. Biol Environ Biol Keys and diagnoses. Part I–Larvae 19:1–457
Environ 101B:65–77 Williams JJ, Brooks SJ, Gosling WD (2012) The response of
Whitlock C, Larsen C (2001) Charcoal as a fire proxy: biological chironomids to Late Pleistocene and Holocene environ-
techniques and indicators. In: Smol JP, Birks HJ, Last WM mental change in the eastern Bolivian Andes. J Paleolim-
(eds) Tracking environmental change using lake sedi- nol. doi:10.1007/s10933-012-9626-1
ments. Kluwer, Dordrecht, pp 75–97
Whitlock C, Dean W, Rosenbaum J, Stevens L, Fritz S, Bracht B,
Power M (2008) A 2650-year-long record of environmental

123