Irruptions of crossbills Loxia spp. in northern Europe –
patterns and correlations with seed production by
key and non-key conifers
RON W. SUMMERS,†,1 BEN SWALLOW,*2 JONAS FRIDMAN,3 TATU HOKKANEN,4 IAN NEWTON5 &
STEPHEN T. BUCKLAND2
1Royal Society for the Protection of Birds, Centre for Conservation Science, Etive House, Beechwood Park,
Inverness, IV2 3BW, UK
2CREEM and School of Mathematics and Statistics, University of St Andrews, St Andrews, KY169LZ, UK
3Swedish University of Agricultural Sciences, Umea˚, 901 83, Sweden
4Natural Resources Institute Finland, Latokartanonkaari 9, 00790, Helsinki, Finland
5Centre for Ecology and Hydrology, Benson Lane, Crowmarsh Gifford, Wallingford, Oxon, UK
Irruptions by boreal seed-eating and frugivorous birds are assumed to be driven by the
production of seeds and fruits, crops of which are highly variable between years. Using
data from Sweden, we tested whether irruptions of Common Crossbills Loxia curvirostra
were correlated with low Norway Spruce Picea abies seed production in the same year as
the irruption and/or high seed production in the year prior to an irruption. Similar tests
were made for Parrot Crossbill Loxia pytyopsittacus irruptions in relation to Scots Pine
Pinus sylvestris seed production. In northern Europe, these conifers represent the key
food species of the two crossbill species, respectively. Despite differing times that seeds
take to mature and asynchronous seed production between the two conifer species,
including a 3-year cycle for Norway Spruce, the two crossbill species often irrupted in
the same year as one another. Analyses showed that irruptions into Britain and other
parts of western Europe by both crossbill species were correlated with low seed produc-
tion by Norway Spruce in Sweden. Low seed production by Scots Pine had a marginally
non-significant additive effect on both crossbill species. In a second set of analyses, the
best-fitting model was one in which low seed production by both conifers in a given year
and high seed production in the previous year were each correlated with large numbers
of irrupting Common and Parrot Crossbills. The models indicate that the incidental
co-occurrence of low seed production of Norway Spruce and Scots Pine in a given year,
after a year of high seed production, may result in an irruption. The seed production of
Norway Spruce and Scots Pine in Sweden was correlated with production by the same
species in Finland, indicating widespread synchrony of cropping across northern Europe.
Keywords: cone-crops, Falsterbo, Finland, seed production, Sweden.
Irruptions of boreal, frugivorous and seed-eating
birds have fascinated ornithologists for decades
(Lack 1954, Sva¨rdson 1957, Newton 1972, 2010,
2023, Alerstam 1990). It is generally assumed that
these movements are related to the reliance that
each bird species has on a particular food, the
availability of which varies synchronously over a
wide area. One hypothesis for synchrony in fruit-
ing is that peaks in production swamp the ability
of frugivorous and seed-eating birds to exploit the
bonanza fully, thereby allowing some seeds a
chance to germinate (Kelly 1994). Years of good
†Present address: Lismore, 7 Mill Crescent, North Kessock,
Inverness, UK
*Corresponding author.
Email: bts3@st-andrews.ac.uk
Twitter: benswallow88
© 2024 The Authors. Ibis published by John Wiley & Sons Ltd on behalf of British Ornithologists' Union.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use,
distribution and reproduction in any medium, provided the original work is properly cited.
Ibis (2024), 166, 1172–1183 doi: 10.1111/ibi.13328
crops are then usually followed by one or more
years with poor crops, a situation that is likely to
cause irruptions.
In northern Europe, Redpolls Acanthis flammea
rely on birch Betula spp. seeds, Thick-billed Nut-
crackers Nucifraga caryocatactes on Hazel Corylus
avellana nuts, Bullfinches Pyrrhula pyrrhula and
Waxwings Bombycilla garrulus on Rowan Sorbus
aucuparia fruits, and crossbills Loxia spp. on coni-
fer seeds (Lack 1954, Newton 1972, 2010, 2023,
Alerstam 1990, Fox et al. 2009). Different bird
species may irrupt together because of synchro-
nous seeding by different food plants (Bock &
Lepthien 1976, Koenig 2001), but others show no
synchrony (Lack 1954).
The hypothesis for the cause of irruptions
revolves around the relationship of the food supply
relative to bird density (Lack 1954, Newton 1970,
2023, Bock & Lepthien 1976, Widrlechner & Dra-
gula 1984). Birds are assumed to irrupt when food
supply is low relative to bird density and this
could be driven by either high bird productivity or
low food supply, or the two could become
entrained if years of low seed/fruit production tend
to follow years of high seed/fruit production.
The species of crossbill in the boreal forests of
continental northern Europe are the Common
Crossbill Loxia curvirostra, Parrot Crossbill Loxia
pytyopsittacus and Two-barred Crossbill Loxia leu-
coptera (Cramp & Perrins 1994). In our study, the
Two-barred Crossbill is not considered because it
is rare in Fennoscandia, as its main breeding range
lies further east. Of the two species considered,
the larger-billed Parrot Crossbill feeds mainly on
seeds of Scots Pine Pinus sylvestris cones, whereas
the smaller-billed Common Crossbill extracts seeds
from the cones of Norway Spruce Picea abies in
northern Europe, though large-billed subspecies of
Common Crossbill that reside in the Mediterra-
nean region forage on Aleppo Pine Pinus halepen-
sis, Mountain Pine Pinus uncinata, Black Pine
Pinus nigra and sometimes Scots Pine (Lack 1944,
Newton 1972, 2023, Senar et al. 1993, Cramp &
Perrins 1994, Summers & Kalejta-Summers 2003).
These differences in diet are consistent with the
idea that different crossbill taxa are adapted to
‘key’ conifer species, although they may also take
seeds from other conifers, including non-native
species (Benkman 1993, Marquiss & Rae 2002,
Summers 2018).
Given the strong associations that Common and
Parrot Crossbills have with Norway Spruce and
Scots Pine, respectively, the irruptions of each
crossbill species are believed to be related to seed
production by the associated conifer species
(Lack 1954, Cramp & Perrins 1994). Features of
the life cycle of the two conifers play an important
role in our understanding of irruptions of the two
crossbill species. The production of Norway
Spruce seed is highly variable, leading to years
with seed abundance (mast years) and other years
with little or no seed production (Hagner 1965,
Broome et al. 2007). The spatial synchrony in
seeding by Norway Spruce can be countrywide
(> 600 km) in Britain, where the species has been
introduced as a commercial crop (Broome
et al. 2007), and across 1000 km in Finland
(Zamorano et al. 2018). Large variability in seed-
ing by Norway Spruce is consistent with the larger
number of Common Crossbill irruptions than Par-
rot Crossbill irruptions (see below) (Newton 1972).
Ringing recoveries and a study of stable isotopes
indicate that irruptions of Common Crossbills into
western Europe originate from different locations
across Europe and western Asia in different years
(Newton 2006, Marquiss et al. 2008).
By contrast, seed production by Scots Pine is
less variable than that of Norway Spruce and
some cones are produced every year
(Hagner 1965, Summers & Proctor 2005, Broome
et al. 2007). Spatial synchrony can be as large as
that of Norway Spruce, varying from 200 km
within Britain (Broome et al. 2007) to 2500 km
within boreal forests (Koenig & Knops 1998).
With a less variable food supply, the Parrot
Crossbill is less prone to irruptions (Newton 1972,
Cramp & Perrins 1994). Moreover, with a smaller
breeding range than that of Common Crossbills,
the origin of their irruptions is restricted to Fen-
noscandia and western Russia (Cramp &
Perrins 1994).
The time taken to form seeds is a further differ-
ence between Norway Spruce and Scots Pine. It
takes 1 year, spread across two calendar years, for
Norway Spruce cones to develop and shed seed
(in spring of the second calendar year), but 2 years
across three calendar years for Scots Pine cones to
do the same (Fletcher 1992). Therefore, the two
conifers are usually out of phase in their seed pro-
duction (Sva¨rdson 1957, Hagner 1965, Broome
et al. 2007), even though cone bud initiation may
be triggered by the same environmental cue, such
as sunlight or warmth (Kosin´ski & Giertych 1982,
Kelly & Sork 2002). On this basis, one would not
© 2024 The Authors. Ibis published by John Wiley & Sons Ltd on behalf of British Ornithologists' Union.
Crossbill irruptions in northwest Europe 1173
 1474919x, 2024, 4, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ibi.13328 by Luonnonvarakeskus, W
iley O
nline Library on [13/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
expect irruptions of Common Crossbills and Parrot
Crossbills to coincide.
Given the hypothesis that Common Crossbill
irruptions are related to Norway Spruce cropping
and Parrot Crossbill irruptions to Scots Pine crop-
ping (Lack 1954, Cramp & Perrins 1994), our
study examined time-series data on seed produc-
tion and crossbill irruptions in northern Europe.
Specifically, we tested the hypothesis that irrup-
tions of Common and Parrot Crossbills are corre-
lated with years of low seed production by
Norway Spruce and Scots Pine, respectively, or/
and with years with large seed production by Nor-
way Spruce and Scots Pine in the year prior to an
irruption, respectively. We also re-examined the
observation that the Common and Parrot Cross-
bills often irrupt in the same years as one another
by extending the period of data to over two centu-
ries (Newton 1972, 2006).
METHODS
Sources of data
Years of crossbill irruptions into regions south of
the boreal zone – into Britain and other parts of
western Europe – were taken from Newton (1972)
for the period 1800–1965, and updated for
1966–2008. These data were acquired from
searches through ornithological journals and, in
Britain, from county avifaunas, county bird reports
and bird observatory reports. All the irruptions
recorded in Britain from the early 20th century
were subsequently also recorded in other western
European countries south of the boreal zone, and
the same is likely to have held for earlier ones. In
the account below, the irruptions are recorded as
occurring in Britain, even though (at least from
the early 20th century) they also extended to
other west European countries.
Annual counts of crossbills were also obtained
from the official migration counts at Falsterbo Bird
Observatory (55.4°N, 12.8°E), at the southern tip
of Sweden, from 1973 to 2018 (Fig. S1). There
are no large conifer forests nearby, indicating that
these birds were on passage. The two species were
distinguished from their flight-calls, and on bill
and head size when birds were close (N. Kjelle´n
pers. comm.). The counts were made from the
same point between 1 August and 20 November
each year by one to two observers operating from
dawn until 14:00 h local time. The starting date of
1 August can miss much of the season for Com-
mon Crossbill movements, which can begin as
early as late June, and are well underway through
July (Newton 1972).
Indices of seed production (cone counts) for
Norway Spruce and Scots Pine were obtained in
five regions aligned north to south in Sweden dur-
ing 1910–1953 from Hagner (1965). These
referred to summer counts of second-summer
cones of Scots Pines and first-summer cones of
Norway Spruce, thereby representing the food
supply for crossbills for the autumn and early win-
ter of the current calendar year through to the late
winter and spring of the following calendar year,
when the seeds are shed. Thus, ‘seed production’
in this paper refers to the seeds available to cross-
bills in the calendar year before the seeds are shed
from cones. Mean values of crop sizes from five
regions were taken as a countrywide Swedish
index. Extending this run of data, additional
annual national data from 1954 to 2020 for the
estimated number of Norway Spruce and Scots
Pine cones were obtained from cone-counting by
the Swedish National Forest Inventory (Fridman
et al. 2014) (Fig. S2).
Data on seed production (seeds per m2) in Fin-
land were obtained from 1960 to 2000 for Norway
Spruces at Heinola (67.8°N, 34.5°E), Kittila¨
(75.5°N, 33.8°E), Kuorevesi (68.8°N, 33.9°E),
Rovaniemi (73.6°N, 34.9°E) and Siilinja¨rvi
(69.9°N, 35.4°E). Similar data were obtained from
1960 to 2004 for Scots Pines at Eckero¨ (67.0°N,
30.9°E), Kittila¨, Kuorevesi, Rovaniemi, Punkaharju
(68.6°N, 36.2°E) and Vippula (68.9°N, 33.7°E).
Means were taken of all sites for each year for
both conifer species. The data on seed production
refer to the calendar year after the year that cone
counts were made in Sweden. To match the Finn-
ish data with the Swedish data and the year in
which crossbills irrupt, the Finnish seed data were
aligned to the previous calendar year (Fig. S3).
Analyses
Comparisons between datasets were limited to
those years with matching data. Indices of seed
production (cone counts) in Sweden were stan-
dardized (setting the mean to 1 and dividing by
the standard deviation). This allowed the two
datasets (1910–1953 and 1954–2020) to be com-
bined. For the 1910–1953 data, synchrony in seed-
ing between the five regions of Sweden was
© 2024 The Authors. Ibis published by John Wiley & Sons Ltd on behalf of British Ornithologists' Union.
1174 R. W. Summers et al.
 1474919x, 2024, 4, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ibi.13328 by Luonnonvarakeskus, W
iley O
nline Library on [13/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
examined using Spearman rank correlations. In
addition, synchrony between the conifers among
regions was examined. Pearson correlation analyses
were used to test for relationships between seed
production of Norway Spruce and Scots Pine
within and between Sweden and Finland. Autocor-
relations tested for regular periodicity in seeding.
Years of crossbill irruptions (Newton 1972 and
updated) were tested against the national seed pro-
duction in Sweden using logistic analyses for
1911–2008. Within binomial models (R Core
Team 2019), years with irruptions were set at 1
and those without were set at 0, and a logit link
function was applied. We tested whether the
probability of an irruption was associated with
indices of seed production by Norway Spruce and
Scots Pine in the current year and the year prior
to an irruption. Interactions were also tested.
Counts of Common and Parrot Crossbills at Fal-
sterbo were each compared with seed production
of Norway Spruce and Scots Pine in Sweden from
1973 to 2018. Hierarchical generalized additive
models (HGAMs) (Pedersen et al. 2019) were
applied, with a crossbill count as the response vari-
able, and the corresponding crossbill species, cone
indices of each of the conifer species and year as
potential explanatory variables. Poisson, Quasi
Poisson and negative binomial were tested as possi-
ble response distributions. Model selection was
based on the lowest Akaike Information Criterion
(AIC) value and/or on explanatory power. These
regression models allowed for both linear and
smooth functions (based on splines) of the covari-
ates to explain variation in crossbill numbers,
including two- and three-way interactions between
conifer species, crossbill species and years. Interac-
tion terms between lagged conifer indices in the
current and previous year were also tested to
check for an association between crossbill count
and the product of the two covariates; for exam-
ple, the product of Norway Spruce in the current
year and Scots Pine in the previous year, testing
for the effect of combined failures of the two
crops or a situation where large crops lead to a
high density of crossbills followed by a shortfall of
food in the following year. The best combination
of interaction terms was chosen to optimize model
fit. The smooth terms were modelled using
thin-plate spline basis functions, except for cross-
bill species, for which a random effect was used.
Terms not deemed to be significant were progres-
sively removed from models and AIC was
recalculated to ensure that both parsimony and
predictive fit were maintained.
The hierarchical structures of the models allow
for varying levels of smoothness to be fitted to
each crossbill species and attributed to
covariate(s), with each smooth term having poten-
tially different levels of smoothness, depending on
the data. Whether they were linear slopes or
smooth effects, (approximate) significance and
hypothesis tests of zero vs. non-zero were calcu-
lated in mgcv (mixed GAM computational vehi-
cle) using methods described in Marra and
Wood (2012) and Wood (2013a, 2013b). After
applying a range of HGAMs to the data in which
the numbers of both crossbill species at Falsterbo
were compared with the Swedish seed production
of both conifers in a given year and/or the previ-
ous year, the best-fitting model was used for
inference.
RESULTS
Crossbill irruptions
From 1800 to 2008 (209 years), 81 Common
Crossbill irruptions and 36 Parrot Crossbill irrup-
tions were documented for Britain and other parts
of western Europe. Although there were fewer
Parrot Crossbill irruptions, many occurred in the
same years as Common Crossbill irruptions, show-
ing a significant positive association (Table 1).
There was also a significant correlation between
the number of Common Crossbills and Parrot
Table 1. The association between irruption years by Common
and Parrot Crossbills from 1800 to 2008. Numbers in paren-
theses are the expected number of years for each irruption
type. Yates’ corrected χ2= 29.9, P< 0.001.
Number of
years with a
Common
Crossbill
irruption
Number of years
without a
Common
Crossbill irruption Total
Number of
years with a
Parrot Crossbill
irruption
29 (14) 7 (22) 36
Number of
years without a
Parrot Crossbill
irruption
52 (67) 121 (106) 173
Total 81 128 209
© 2024 The Authors. Ibis published by John Wiley & Sons Ltd on behalf of British Ornithologists' Union.
Crossbill irruptions in northwest Europe 1175
 1474919x, 2024, 4, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ibi.13328 by Luonnonvarakeskus, W
iley O
nline Library on [13/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Crossbills counted each year at Falsterbo (Fig. 1;
Fig. S1). The median number of Common Cross-
bills, 508 (range 0–31 626), was similar to that for
Parrot Crossbills, 439 (range 0–4490) (Mann–
Whitney U= 1150.5, P= 0.47).
To check whether both datasets on crossbills
(from Britain and Falsterbo) were describing the
same pattern of irruptions, the 36 years
(1973–2008) that provided data for both locations
were compared. For Common Crossbills, which
had 10 irruptions in this period, the mean counts
of crossbills at Falsterbo in years classified as irrup-
tion and non-irruption years in Britain were 5017
and 846, respectively, a difference that is statisti-
cally significant (Mann–Whitney U = 60,
P= 0.013). For Parrot Crossbills, which had six
irruptions, the mean counts were 1023 and 611,
respectively, a difference that was not statistically
significant (Mann–Whitney U = 51, P= 0.098) but
followed the same pattern as the Common
Crossbills.
Seed (cone) production in Sweden
For 1909–1953, there was significant synchrony
between neighbouring regions for the indices of
seed production for each of the conifers, and par-
ticularly for Norway Spruce (Table 2). The degree
of synchrony between regions declined with
increasing distance between pairs of regions,
though all coefficients were positive.
There were no significant correlations between
Norway Spruce and Scots Pine indices of seed pro-
duction either nationally (r=
0.044, n= 109,
n.s.) or for any of the five regions (Table 2,
Fig. S2).
Auto-correlations for Norway Spruce seed pro-
duction revealed a significant negative lag at year 1
for three regions (regions 2, 3 and 5) and for the
national indices (Fig. 2), showing that years of low
seed production tended to follow years of high
production. There was also a positive correlation
at year 3 in regions 4 and 5, and nationally
(Fig. 2), indicating a 3-year cycle in seed produc-
tion. For Scots Pine, no significant
auto-correlations were detected (Fig. 2).
Seed production in Finland
There was no significant correlation between Nor-
way Spruce and Scots Pine in seed production
(r= 0.026, n= 41, n.s.) (Fig. S3). There was a sig-
nificant increase over time in seed production of
Norway Spruce (rs= 0.467, n= 41, P< 0.005) but
no trend for Scots Pine (rs= 0.219, n= 45, ns).
Auto-correlation analyses showed a weak effect
for a 3-year cycle in Norway Spruce seed produc-
tion, but there was no evidence for regular cycling
in Scots Pine seed production (Fig. 2).
For each of the conifer species, seed production
in Finland and Sweden were significantly corre-
lated (Fig. 3).
The relationship between crossbill
irruptions and seed production
Years of Common and Parrot Crossbill irruptions
were correlated with low seed production of Nor-
way Spruce in Sweden. There were additional
marginally non-significant indications that irrup-
tions could be related to low seed production by
Scots Pine, but there were no significant non-linear
interactions between the two conifers (Table 3),
nor were there significant associations with the
Norway Spruce and Scots Pine seed crops in the
year prior to irruptions, either in association with
the current year’s seed production or
independently.
The final GAM resulted in the predicted cross-
bill count as below, following backwards elimina-
tion of non-significant covariates:
E (crossbill count) = exp (slope (Norway
Spruce)+ slope (Scots Pine lagged)+ slope
Figure 1. The relationship between the log number (+1) of
Parrot Crossbills against log number (+1) of Common Cross-
bills at Falsterbo, southern Sweden, from 1973 to 2018
(r = 0.898, n= 46, P< 0.001).
© 2024 The Authors. Ibis published by John Wiley & Sons Ltd on behalf of British Ornithologists' Union.
1176 R. W. Summers et al.
 1474919x, 2024, 4, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ibi.13328 by Luonnonvarakeskus, W
iley O
nline Library on [13/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
(Norway Spruce:Scots Pine lagged)+ smooth
(Scots Pine) + smooth (Norway Spruce) + smooth
(Scots Pine lagged)+ smooth (Norway Spruce
lagged)+ smooth (year) + smooth (crossbill spe-
cies)), where the crossbill count followed a nega-
tive binomial distribution. Slope refers to a linear
trend, consisting of a constant multiplicative rate
of change, whereas a smooth term is made up of a
spline basis function. Models are fitted using the
gam function in the R package mgcv.
In this model, large numbers of crossbills were
related to low seed production of both conifers in
a given year, plus combined high seed production
of both conifers in the previous year (Table 4).
Figure 4 shows the relationship between numbers
of both species of crossbill passing through Fal-
sterbo and the corresponding seed indices of the
given year and previous year. The smooth relation-
ship with the current year suggests that both cross-
bill species are likely to irrupt at combined lower
levels of seed production. The estimate of the neg-
ative binomial variance scale parameter θ was
0.59, suggesting greater variance relative to the
equivalent Poisson distribution. For the negative
binomial, the variance is Var yð Þ ¼ μþ μ2θ , whereas
for a Poisson model it is Var yð Þ ¼ μ.
DISCUSSION
The positive correlation between the irruptions of
Common and Parrot Crossbills, as observed by
Newton (1972), is supported statistically. In addi-
tion, counts at Falsterbo confirmed that numbers
of both species can fluctuate together, despite
being associated with different key conifers that
have different life cycles and show no general
synchrony in their seed production, either in Swe-
den or Finland. For the years in which we had data
on irruptions in Britain and counts in Falsterbo, a
significant positive correlation emerged for the
Common Crossbill. For Parrot Crossbills, which
had fewer irruptions, there was no significant cor-
relation but a similar pattern. Numbers of the two
crossbill species counted at Falsterbo were similar
to one another, although this could be an artefact
of the start date of counting, which will have
missed the early part of Common Crossbill irrup-
tions (Newton 1972). The reason why the two
crossbills irrupt at different dates within a year is
unclear.
The variation in seeding followed patterns that
have been described previously: greater
year-to-year variation in the seeding of Norway
Spruce than of Scots Pine, with peaks in cone pro-
duction every few years (Hagner 1965, Broome
et al. 2007). Reinikainen (1937) noted that Nor-
way Spruce has peaks in seeding every 3–5 years in
Finland, while Newton (1972) listed 2-year seed-
ing for the Alps, 3–4 years in the southern boreal
region and 5 years in the north. Broome
et al. (2007) recorded a 4-year pattern in the UK.
Our study noted that for the Norway Spruce, seed
production followed a 3-year cycle, suggesting
some intrinsic governing factor, rather than an
extrinsic factor such as the weather (Zamorano
et al. 2018). Although there was no evidence of
cycling by Scots Pines in this study, there is some
evidence for a 3-year cycle in Scotland (Sum-
mers 2011). However, there was no indication of
regular cycling in the irruptions of crossbills.
The tendency for Norway Spruce to have a
poor seed (cone) crop following a large crop has
Table 2. Spearman correlation coefficients for cone indices comparing different conifers and pairs of five regions of Sweden.
SP1 SP2 SP3 SP4 SP5 NS1 NS2 NS3 NS4
SP2 0.79
SP3 0.63 0.73
SP4 0.17 0.17 0.29
SP5 0.14 0.27 0.28 0.51
NS1 
0.03 0.01 
0.07 
0.06 
0.14
NS2 0.02 0.09 0.10 0.24 0.08 0.77
NS3 
0.12 
0.10 0.00 0.21 
0.05 0.62 0.80
NS4 
0.07 
0.19 
0.17 0.30 0.02 0.32 0.49 0.78
NS5 
0.30 
0.29 
0.25 0.24 0.01 0.33 0.45 0.70 0.91
NS, Norway Spruce; SP, Scots Pine. 1 is the most northerly region and 5 the most southerly. n= 45. Those coefficients which indi-
cate strong correlations (above 0.7) are in bold.
© 2024 The Authors. Ibis published by John Wiley & Sons Ltd on behalf of British Ornithologists' Union.
Crossbill irruptions in northwest Europe 1177
 1474919x, 2024, 4, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ibi.13328 by Luonnonvarakeskus, W
iley O
nline Library on [13/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
previously been noted by Tiren (1935) and for
Scots Pine by Watson et al. (2009). For Norway
Spruce, poor cropping may be influenced not only
by the low nutrient stores in the trees, but also by
insect and rust fungus species that can damage a
large part of a seed crop, especially after years
with high cone production (Annila 1981,
Kaitera 2013).
The positive correlation in Sweden and between
Sweden and Finland in seeding by Norway
Spruces, and the same pattern occurring separately
for Scots Pines, showed that synchrony in seeding
within each conifer species was widespread. We
obtained no data from Russia to test for correla-
tions with conifers further east, but by predicting
cone crops from tree ring data, Koenig and
Knops (1998) suggested that seed crops can be
synchronized across 500 km for Norway Spruce
and 2500 km for Scots Pines. According to Zamor-
ano et al. (2018), Norway Spruce showed spatial
Figure 2. Auto-correlation coefficients at different time lags for national indices for Norway Spruce and Scots Pine cone (seed) pro-
duction in Sweden from 1910 to 2020, and seed indices for Finland from 1960 to 2000 for Norway Spruce and 1960 to 2004 for
Scots Pine. The dashed lines show the critical values where P= 0.05.
© 2024 The Authors. Ibis published by John Wiley & Sons Ltd on behalf of British Ornithologists' Union.
1178 R. W. Summers et al.
 1474919x, 2024, 4, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ibi.13328 by Luonnonvarakeskus, W
iley O
nline Library on [13/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
synchrony in seed production at scales up to
1000 km. Thus, the correlation in seed production
within species, as seen between Sweden and Fin-
land, may extend well into Russia.
Joint irruptions by Common and Parrot Cross-
bills indicated that they are not independently gov-
erned by the seed production of their key conifer
(Fig. 4). Both sets of analyses showed an interplay
between both crossbill species and both conifer
species. Low seed production, especially for Nor-
way Spruce, had an apparent influence on the
irruptions of both crossbills (first analysis, Table 3),
and the second analysis revealed that counts of
both crossbill species were related to seed produc-
tion by both conifers in the year of an irruption
and the year prior to an irruption. The initial posi-
tive correlation between crossbill counts and cone
crop size (Fig. 4) may suggest that irruptions are
Figure 3. Relationships between standardized indices of seed
(cone) production for Norway Spruce (1961–1999) and Scots
Pine (1960–2003) in Sweden and Finland in matching years.
r= 0.63, n= 39, P< 0.001 for Norway Spruce and r = 0.51,
n= 44, P< 0.001 for Scots Pine.
Table 3. Results of logistic regression analyses of crossbill
irruptions in relation to standardized cone (seed) indices of
abundance in the current year in Sweden from 1911 to 2008.
df Estimate (se) Z (P )
Common Crossbill irruptions
Intercept 1 
0.962 (0.260) 
3.70 (< 0.0001)
Norway Spruce
index
1 
0.813 (0.305) 
2.67 (0.0077)
Scots Pine index 1 
0.450 (0.246) 
1.83 (0.067)
Dispersion
parameter* = 1.17
Parrot Crossbill irruptions
Intercept 1 
2.142 (0.484) 
4.42 (< 0.0001)
Norway Spruce
cone index
1 
1.706 (0.598) 
2.85 (0.0044)
Scots Pine cone
index
1 
0.583 (0.305) 
1.91 (0.056)
Dispersion
parameter* = 0.86
*The dispersion parameter would be 1 if there was no under-
or overdispersion.
Table 4. Summary statistics of a negative binomial HGAM
fitted to Falsterbo crossbill counts (1973–2018) in relation to
seed production of Scots Pine and Norway Spruce in the cur-
rent year and with lags.
Parameter
Estimate
(se) Z/χ2 P
Model (crossbill[t]=Norway Spruce[t] × Scots Pine[t]) + (spruce
[t–1] × Scots Pine[t–1])
Scots Pine 0.27
(0.020)

4.574 <0.0001
Norway Spruce 0.35
(0.044)
8.324 <0.0001
Pine × spruce 0.004
(0.001)
3.095 0.002
Smooth (year) – 15.351 0.001
Smooth (Scots Pine) – 30.206 <0.0001
Smooth (Norway Spruce) – 36.572 <0.0001
Smooth (Scots Pine lag) – 79.470 <0.0001
Smooth (Norway Spruce
lag)
– 12.778 0.003
Smooth (crossbill
species)
– 4.556 0.017
Significant fixed and smooth effects are shown. Z-scores are
for the fixed effects and chi-square values are for the smooth
effects.
© 2024 The Authors. Ibis published by John Wiley & Sons Ltd on behalf of British Ornithologists' Union.
Crossbill irruptions in northwest Europe 1179
 1474919x, 2024, 4, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ibi.13328 by Luonnonvarakeskus, W
iley O
nline Library on [13/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
likely to occur prior to the complete failure of a
crop. This is supported by the interaction plots
that show that the largest irruptions tend to occur
when low (but not complete failure of) Norway
Spruce seed availability is combined with reason-
ably low (middle tercile) Scots Pine seed availabil-
ity. In other words, it is the lack of an alternative
seed crop when the seed crop of the key conifer is
low that drives irruptions. The second peak in the
interaction plots corresponds to very high seed
indices of both Norway Spruce and Scots Pine and
may suggest that years with very high seed crops
attract crossbills to feed on the available crop and
to breed successfully.
The fact that irruptions by both crossbill species
showed relationships with seed indices in Sweden
may seem surprising because irrupting crossbills do
not come from just Fennoscandia. There have
been ringing recoveries of Common Crossbills
from as far as central Russia (Newton 2006), and
isotope data from feathers of irruptive birds imply
that some irruptions begin even further east (Mar-
quiss et al. 2008, 2012). However, if those birds
irrupting from regions east of Sweden move west
through the boreal forest, they are likely to arrive
in Sweden. This is the last country with a substan-
tial amount of conifer forest, so any onward move-
ment into regions further south could be
Figure 4. Fitted smooth relationships between crossbill numbers at Falsterbo and Norway Spruce and Scots Pine seed indices in
the current (top) and previous (bottom) year, as estimated from the HGAM (Table 4). In each subfigure, the two left-hand plots show
the fitted smooth relationships between Common (dark lines) and Parrot Crossbills (light lines) and Norway Spruce and Scots Pine
indices, with 95% confidence regions in the corresponding dashed lines. The right-hand two interaction plots show the estimated
marginal relationships between crossbill numbers and the Norway Spruce seed index, for the cases where the corresponding Scots
Pine seed index was low, medium or high.
© 2024 The Authors. Ibis published by John Wiley & Sons Ltd on behalf of British Ornithologists' Union.
1180 R. W. Summers et al.
 1474919x, 2024, 4, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ibi.13328 by Luonnonvarakeskus, W
iley O
nline Library on [13/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
influenced by the seed crop in Sweden. It is less
surprising that Parrot Crossbill irruptions are influ-
enced by the seed crop in Sweden because their
breeding range is restricted to Fennoscandia and
western Russia (Cramp & Perrins 1994).
Although the two crossbill species are associ-
ated with different key conifers as food sources,
two factors may account for the correlations we
found. First, seed crops for a given conifer fluctu-
ate in synchrony over wide areas, exemplified by
the correlation in seeding between Sweden and
Finland, and tree ring data (a surrogate for seed-
ing) indicate synchrony over 2500 km for Scots
Pine (Koenig & Knops 1998). Secondly, although
Parrot Crossbills forage largely on Scots Pine, they
also utilize Norway Spruce, Sitka Spruce Picea
sitchensis and larch Larix spp. (Cramp & Per-
rins 1994, Marquiss & Rae 2002). Similarly, Com-
mon Crossbills make use of several species of
conifers: Sitka Spruce, Lodgepole Pine Pinus con-
torta (both are North American conifers intro-
duced to Europe for timber cropping), larch and
Scots Pine (Marquiss & Rae 2002, Summers
et al. 2002). They also breed in pure Scots Pine
forests when the cones open to release seeds in
spring (Haapanen 1966, Summers & Proctor 2005,
Summers et al. 2010). Therefore, crossbills are
flexible in their use of conifers. It would seem that
if, having bred successfully in a year when seed
production by both conifers is high, and there is
low seed production by both conifers in the fol-
lowing year, there is no alternative for the cross-
bills but to move outside their normal ranges in
search of food. Such combinations in seed produc-
tion by both conifers will occur only by chance,
given that seeding is normally asynchronous.
Nils Kjelle´n kindly provided the counts of crossbills at
Falsterbo, Sweden. Jeremy Wilson commented on drafts
of this paper and the paper was reviewed by Craig
Benkman, Maggie MacPherson and an anonymous
reviewer.
AUTHOR CONTRIBUTIONS
Ron W. Summers: Conceptualization; investigation;
data curation; formal analysis; methodology; writing
– original draft; writing – review and editing. Ben T.
Swallow: Data curation; investigation; formal analy-
sis; methodology; writing – original draft; writing –
review and editing. Jonas Fridman: Data curation;
investigation; writing – review and editing. Tatu
Hokkanen: Data curation; writing – review and
editing; investigation. Ian Newton: Data curation;
investigation; writing – original draft; writing –
review and editing. Stephen T. Buckland: Formal
analysis; writing – review and editing.
FUNDING
None.
CONFLICT OF INTEREST
STATEMENT
The authors have no conflicting interests to
declare.
ETHICAL NOTE
None.
CONSENT FOR PUBLICATION
Not applicable.
Data Availability Statement
All data required to run the analysis will be made
available by the authors on reasonable request.
REFERENCES
Alerstam, T. 1990. Bird Migration. Cambridge: Cambridge
University Press.
Annila, E. 1981. Kuusen ka¨py- ja siementuholaisten
kannanvaihtelu. Summary: fluctuations in cone and seed
insect populations in Norway Spruce. Commun. Inst. For.
Fenn. 101: 1–32.
Benkman, C.W. 1993. Adaptation to single resources and the
evolution of crossbill (Loxia) diversity. Ecol. Monogr. 63:
305–325.
Bock, C.E. & Lepthien, L.W. 1976. Synchronous eruptions of
boreal seed-eating birds. Am. Nat. 110: 559–571.
Broome, A., Hendry, S. & Peace, A. 2007. Annual and
spatial variation in coning shown by the forest condition
monitoring programme data for Norway spruce, Sitka spruce
and Scots pine in Britain. Forestry 80: 17–28.
Cramp, S. & Perrins, C.M. (eds) 1994. The Birds of the
Western Palearctic, Vol. 8. Oxford: Oxford University Press.
Fletcher, A.M. 1992. Flower, fruit and seed development and
morphology. For. Commission Bull. 83: 59–70.
Fox, A.D., Kobro, S., Lehikoinen, A., Lyngs, P. & Va¨isa¨nen,
R.A. 2009. Northern Bullfinch Pyrrhula p. pyrrhula irruptive
behaviour linked to rowanberry Sorbus aucuparia
abundance. Ornis Fennica 86: 51–60.
© 2024 The Authors. Ibis published by John Wiley & Sons Ltd on behalf of British Ornithologists' Union.
Crossbill irruptions in northwest Europe 1181
 1474919x, 2024, 4, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ibi.13328 by Luonnonvarakeskus, W
iley O
nline Library on [13/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Fridman, J., Holm, S., Nilsson, M., Nilsson, P., Ringvall,
A.H. & Sta˚hl, G. 2014. Adapting National Forest Inventories
to changing requirements – the case of the Swedish
National Forest Inventory at the turn of the 20th century.
Silva Fenn. 48: 1095.
Haapanen, A. 1966. Bird fauna of the Finnish forests in
relation to forest succession. Ann. Zool. Fenn. 3: 176–200.
Hagner, S. 1965. Cone crop fluctuations in Scots pine and
Norway spruce. Stud. Forestal. Suec. 33: 1–21.
Kaitera, J. 2013. Thekopsora and Chrysomyxa cone rusts
damage Norway spruce cones after a good cone crop in
Finland. Scand. J. Forest Res. 28: 217–222.
Kelly, D. 1994. The evolutionary ecology of mast seeding.
Trends Ecol. Evol. 9: 465–470.
Kelly, D. & Sork, V.L. 2002. Mast seeding in perennial plants:
why, how, where? Annu. Rev. Ecol. Syst. 33: 427–447.
Koenig, W.D. 2001. Synchrony and periodicity of eruptions by
boreal birds. Condor 103: 725–735.
Koenig, W.D. & Knops, J.M.H. 1998. Scale of mast-seeding
and tree-ring growth. Nature 396: 225–226.
Kosin´ski, G. & Giertych, M. 1982. Light conditions
inside developing buds affect floral induction. Planta 155:
93–94.
Lack, D. 1944. Correlation between beak and food in the
Crossbill, Loxia curvirostra Linnæus. Ibis 86: 522–553.
Lack, D. 1954. The Natural Regulation of Animal Numbers.
Oxford: Clarendon Press.
Marquiss, M. & Rae, R. 2002. Ecological differentiation in
relation to bill size amongst sympatric, genetically
undifferentiated crossbills Loxia spp. Ibis 144: 494–508.
Marquiss, M., Hobson, K.A. & Newton, I. 2008. Stable
isotope evidence for different regional source areas of
Common Crossbill Loxia curvirostra irruptions into Britain. J.
Avian Biol. 39: 30–34.
Marquiss, M., Newton, I., Hobson, K.A. & Kolbeinsson, Y.
2012. Origins of irruptive migrations by Common Crossbills
Loxia curvirostra into northwestern Europe revealed by
stable isotope analysis. Ibis 154: 400–409.
Marra, G. & Wood, S.N. 2012. Coverage properties of
confidence intervals for generalized additive model
components. Scand. J. Stat. 39: 53–74.
Newton, I. 1970. Irruptions of crossbills in Europe. In Watson,
A. (ed) Animal Populations in Relation to their Food
Resources: 337–357. Oxford: Blackwell.
Newton, I. 1972. Finches. London: Collins.
Newton, I. 2006. Movement patterns of Common Crossbills
Loxia curvirostra in Europe. Ibis 148: 782–788.
Newton, I. 2010. Bird Migration. London: Collins.
Newton, I. 2023. The Ecology of Bird Migration, 2nd edn.
London: Academic Press.
Pedersen, E.J., Miller, D.L., Simpson, G.L. & Ross, N.
2019. Hierarchical generalized additive models in ecology:
an introduction with mgcv. PeerJ 7: e6876.
R Core Team 2019. R: A language and environment for
statistical computing. Vienna: R Foundation for Statistical
Computing. https://www.R-project.org/
Reinikainen, A. 1937. The irregular migrations of the
Crossbill, Loxia c. curvirostra, and their relation to the
cone-crop of the conifers. Ornis Fennica 14: 55–64.
Senar, J.C., Borras, A., Cabrera, T. & Cabrera, J. 1993.
Testing for the relationship between coniferous crop stability
and Common Crossbill residence. J. Field Ornithol. 64:
464–469.
Summers, R.W. 2011. Patterns of exploitation of annually
varying Pinus sylvestris cone crops by seed-eaters of
differing dispersal ability. Ecography 34: 723–728.
Summers, R.W. 2018. Foraging patterns of Common
Crossbills (Loxia curvirostra) on spruces (Picea spp.) in
Scotland. Forestry 91: 444–450.
Summers, R. & Kalejta-Summers, B. 2003. Preliminary
observations on the feeding habits and habitats of crossbills
Loxia curvirostra balearica at Cape Formentor, Majorca.
Anu. Ornitol. Balear. 18: 3–15.
Summers, R.W. & Proctor, R. 2005. Timing of shedding
seeds and cones, and production in different stands of
Scots pines at Abernethy Forest, Scotland. Forestry 78:
541–549.
Summers, R.W., Jardine, D.C., Marquiss, M. & Rae, R.
2002. The distribution and habitats of crossbills Loxia spp. in
Britain, with special reference to the Scottish Crossbill Loxia
scotica. Ibis 144: 393–410.
Summers, R.W., Dawson, R.J.G. & Proctor, R. 2010.
Temporal variation in breeding and cone size selection by
three species of crossbills Loxia spp. in a native Scots
pinewood. J. Avian Biol. 41: 219–228.
Sva¨rdson, G. 1957. The ‘invasion’ type of bird migration. Br.
Birds 50: 314–343.
Tiren, L. 1935. Om granens kottsa¨ttning, dess periodicitet och
samband med temperatur och nederbo¨rd. Summary: on the
fruit setting of spruce, its periodicity and relation to
temperature and precipitation. Medd. Stat. Skogsf. Anst. 28:
413–415.
Watson, A., Marquiss, M. & Summers, R. 2009. Abundance
of crossbills, siskins and cone-crops. Ornis Fennica 86:
38–40.
Widrlechner, M.P. & Dragula, S.K. 1984. Relation of
cone-crop size to irruptions of four seed-eating birds in
California. Am. Birds 38: 840–846.
Wood, S.N. 2013a. A simple test for random effects in
regression models. Biometrika 100: 1005–1010.
Wood, S.N. 2013b. On p-values for smooth components of an
extended generalized additive model. Biometrika 100:
221–228.
Zamorano, J.G., Hokkanen, T. & Lehikoinen, A. 2018.
Climate-driven synchrony in seed production of masting
deciduous and conifer tree species. J. Plant Ecol. 11:
180–188.
Received 18 September 2023;
Revision 29 February 2024;
revision accepted 2 April 2024.
Associate Editor: Maggie MacPherson.
SUPPORTING INFORMATION
Additional supporting information may be found
online in the Supporting Information section at
the end of the article.
Figure S1. Standardized numbers of crossbills at
Falsterbo, southern Sweden, from 1973 to 2018.
© 2024 The Authors. Ibis published by John Wiley & Sons Ltd on behalf of British Ornithologists' Union.
1182 R. W. Summers et al.
 1474919x, 2024, 4, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ibi.13328 by Luonnonvarakeskus, W
iley O
nline Library on [13/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Figure S2. Standardized values of Norway
Spruce and Scots Pine cone (seed) production in
Sweden from 1910 to 1953 (Hagner 1965) and
1954 to 2020 (Fridman et al. 2014).
Figure S3. Standardized values for Norway
Spruce and Scots Pine seed production in Finland
from 1960 to 2000 for Norway Spruce and 1960
to 2004 for Scots Pine.
© 2024 The Authors. Ibis published by John Wiley & Sons Ltd on behalf of British Ornithologists' Union.
Crossbill irruptions in northwest Europe 1183
 1474919x, 2024, 4, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ibi.13328 by Luonnonvarakeskus, W
iley O
nline Library on [13/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License