Jökull - 01.12.2007, Side 34

Maciej Dąbski
DISCUSSION
Bradwell (2004) and McKinzey et al. (2004) obta-
ined lichenometric frequency gradients for moraines
in SE Iceland using all thalli (also the largest ones)
within each population. However, for one moraine at
Hólárjökull, Bradwell (2004) obtained inaccurate age
estimations due to the single largest lichen which sho-
uld be, in his opinion, omitted. If the largest thalli
at Fláajökull are included in the analysis, the obta-
ined ages are obviously incorrect, since moraine III
cannot be older then moraine II (Figures 3). Rela-
tively good correlation between previously obtained
dates (Dąbski, 2002) for Fláajökull moraine ridges I–
IV and the lichenometric dates obtained after omitting
the largest thalli shows the advantage of the described
modification of input data into the lichenometric size-
frequency method. A slight but steady decrease in the
value of R2 in the direction of the oldest moraine is lo-
gical, and can be explained by the growing age of the
lichen populations. Bradwell (2004) andMcKinzey et
al. (2004) found a similar regularity in R2 value chan-
ges while surveying moraines of Fjallsjökull, Hólár-
jökull, Skálafellsjökull and Heinabergsjökull, neigh-
bours of Fláajökull (Figure 1). It must be emphasi-
sed that if the largest thalli growing on Fláajökull mo-
raines were included into the lichenometrical dating,
there would not be a gradual and logical change in the
frequency gradient and R2 value. The method used
in this research is sound and based on the assumption
that the population frequency gradient should be cal-
culated only by omitting exceptionally large thalli that
constitute the right outliers of the frequency distribu-
tion graphs (not only the single largest thallus). Such
a procedure produces the most reasonable dates for
Fláajökull moraine ridges (Table 2).
The lichen size-frequency gradient method used
on moraine ridge V yields a date (AD 1944), which
is 9-10 years younger than the date derived in pre-
vious research (Dąbski, 2002), Table 2. This discre-
pancy cannot be attributed to the described reduc-
tion method since there are no exceptionally large li-
chens in this population and the method was not im-
plied on this ridge (Figure 3). Superficial material
on ridge V was deposited in 1934/35 according to
Dąbski (2002). This date is based on direct glacio-
logical measurements (Sigurdsson, 2000) performed
along the Hólmsárgarður profile situated in the south-
western part of the Fláajökull moraines (Figure 1).
Therefore, the date refers precisely to the run of the
profile. The measurements revealed a transgression
in 1941 of 59 m along the profile followed by further
regression. Taking under consideration that 1) the te-
sting field for lichenometrical measurements covered
300 m of the ridge length, 2) the fact that Fláajökull
glacier front has many small-scale lobes, and 3) the
mentioned short-term transgression, the several-year
difference between glaciological measurement along
the profile and the lichenometrical date for the whole
testing field can be accepted.
Dates derived with the use of Bradwell’s (2001)
dating curve based on a single largest lichen (without
the single exceptionally large thallus) suggest that the
Fláajökull moraine ridges I, II, III and IV were pro-
duced in the 19th century (Table 2) which is, in light
of other data (Dąbski, 2002), impossible. Extrapola-
tion of the frequency gradient trend line allowed to
obtain the “1-in-1000” lichen thallus for each popu-
lation, Table 2 (Locke et al. 1979). Dates calculated
with the use of such thalli and the Bradwell’s (2001)
dating curve show that the oldest Fláajökull end mo-
raine was abandoned in the beginning of the 19th cen-
tury, which would support the findings of McKinzey
et al. (2004). Nevertheless, these dates cannot be ac-
cepted because they disagree with the direct glacio-
logical measurements, which is the case for moraine
ridge V (Figure 1, Table 2).
Discrepancies between Evans et al. (1999) ap-
proach, based on the average of five largest thalli,
the dates obtained in this study and the former da-
ting (Dąbski, 2002), Table 2, result from the diffe-
rent lichenometrical method implied. Possible insta-
bility of the moraine surface due to differential mel-
ting of buried glacial ice (Everest and Bradwell, 2003)
can add to the differences obtained in dating, altho-
ugh the buried ice was found only within the youn-
gest moraine ridge. Other obstacles for use of liche-
nometry in dating moraine ridges of Fláajökull can
stem from uncertainty of colonisation lag time, un-
32 JÖKULL No. 57