Content uploaded by Diego Rubolini
Author content
All content in this area was uploaded by Diego Rubolini on Mar 05, 2019
Content may be subject to copyright.
Folia Zool. – 52(1): 67–76 (2003)
Influence of seasonality, temperature and rainfall on the winter diet
of the long-eared owl, Asio otus
Diego RUBOLINI1*, Andrea PIROVANO2and Sergio BORGHI3
1Dipartimento di Biologia Animale, Universitàdegli Studi di Pavia, p.zza Botta 9, I-27100 Pavia, Italy;
e-mail: diego.rubolini@unipv.it
2Wildlife Biology and Management Unit, Center for Life Sciences Weihenstephan, Technische Universität
München, Am Hochanger 13, D-85354 Freising, Germany
3Osservatorio Meteorologico di Milano-Duomo, piazza Duomo 21, I-20121 Milano, Italy
Received 22 January 2002; Accepted 16 August 2002
Abstract. In this study we analysed whether the diet composition of a wintering population
(40–70 individuals) of long-eared owls (Asio otus) in northern Italy showed within-season
variation, and whether it was influenced by ambient temperature and rainfall. Diet composition
was determined by pellet content, and over 5500 prey items were analysed; pellets were
collected at 2-wk intervals over two consecutive winters (October to April), 1996–1998. Three
out of five main prey categories showed a marked within-season variability in relative frequency
in diet, both considering the number of prey items and prey biomass, whereas between-year
variability was shown only by a single prey category (Savi’s pine vole). Although rainfall had no
influence on diet composition, temperature affected negatively the prevalence of harvest mouse,
a relatively unimportant prey category. Thus, the considered weather variables seem to have little
influence on the winter diet composition (at the level of individual prey categories) of these owls.
However, diet breadth (estimated by the Levins’index of niche breadth) increased with increasing
rainfall and decreasing temperature, when calculated on the proportion of prey items: hence it
seems that the owls become more generalists under unfavourable weather conditions.
Key words: weather conditions, nocturnal raptors, small mammals, niche breadth, seasonal variation
Introduction
Meteorological conditions can affect the diet and hunting success of raptors. For example,
rain seems to have a negative impact on the hunting performance of some species, hampering
flight or perceptive ability (Barbieri et al. 1975, W ijnandts 1984, Michelat &
Giraudoux 1992, Olsen & Olsen 1992, Henrioux 1999). In the long-eared
owl (Asio otus Linnaeus, 1758), a low ambient temperature and presence of rain seems to
reduce flying activity, although the effect varied according to season (Henrioux 1999).
Also, surface activity patterns of small mammals (the commonest prey of the long-eared owl;
Marti 1976) are influenced by the weather to varying degrees (e.g. S i dorowicz
1960, Getz 1961, M aguire 1999). Further, diet composition may also depend on
seasonal variation (hereafter defined as seasonality) in prey choice by the owls and/or prey
activity, as shown by previous analyses of the owl’s diet (Nilsson 1981, P irovano et
al. 2000a). The most comprehensive study so far is that of Nilsson (1981) in southern
Sweden, which investigated changes in diet composition througout several year cycles.
However, the data were grouped by month, and were mostly discussed in terms of between-
year and between-season variations in prey and habitat choice, whereas no effort has been
devoted to the analysis of within-season variability patterns in diet composition.
67
*Corresponding author
To our knowledge, few studies addressed whether there are relationships between
weather conditions and diet composition in free-living raptors, while accounting for
seasonal variation in diet (e.g. Canova 1989). Here we analysed within-season variation
in diet composition (determined by pellet content) in a population of long-eared owls
studied over two consecutive winter seasons (October to April), 1996–1998. The diet and
behaviour of this urban wintering population has been described elsewhere (P irovano
et al. 2000a,b): the brown rat (Rattus norvegicus) was the most abundant prey in terms of
biomass, and the monthly proportion of rats in the diet was inversely related to mean
monthly rat weight (P irovano et al
.2000a), suggesting a selection on young rats
(90–100 g). The aim of the present study was to assess the effect of seasonality on diet and
the relationships between diet composition and meteorological variables (mean temperature
and amount of rainfall preceding pellet collection) and determine which of these factors
explained the largest variation in diet.
Material and Methods
The roost site, occupied by 40–70 owls during winter (October to April), was located in the
southern tip of the city of Milan (northern Italy, 45°28’N – 9°12’E). Owls hunt in the suburbs
and cultivated fields surrounding the city (P irovano et al. 1997, 2000a,b). The owls’
likely hunting range consisted mainly of winter stubbles (cereal crops, colza), poplar
plantations, meadows, together with patches of network habitats (copses, hedgerows along
field margins), which are actively selected by owls whilst searching for prey in our study
region (Galeotti et al. 1997). Mean monthly rainfall in the study period was 80 mm per
month (ranging from 150 mm in December to 6 mm in March), while mean monthly
temperatures ranged from 14.6°C in October to 3.9°C in January. The study was carried out
over two consecutive winters (November–April 1996–97 and October–April 1997–98).
Pellets were collected at 2 wk-intervals each month (first collection: 15th of each month;
second collection: 30th/31st) and analysed following standard techniques (Y alden &
Norris 1990). Possible biases associated with pellet collection were minimised by
collecting all the intact pellets at each visit, always along the same trail below the owls’
perches. Apreliminary cleaning-up of all pellet remains was undertaken 2 wk before the first
collection for each of the two study years. As in other studies, mammals were determined to
species level, and birds were considered as a single category (Galeotti & Canova
1994, P irovano et al. 2000a). Weather variables were recorded at Brera-Duomo
Meteorological Observatory (3 km from the roost site) at hourly intervals. In the analyses, we
concentrated only on main prey categories (those constituting at least 5% of prey items over
the two years, following the definition given in P irovano et al. 2000a): they were wood
mouse (Apodemus sylvaticus) (38.5 %), brown rat (21.7 %), Savi’s pine vole (Microtus savii)
(17.3 %), birds (9.6 %), harvest mouse (Micromys minutus) (5.5 %). Main prey categories
made up 93 % of prey items over the two years. The complete dataset used in the analyses is
reported in the Appendix.
Statistical analyses
Using mean hourly temperature data, we calculated the daily mean temperatures for 2-week
periods, and rainfall was measured as the amount of rain (in mm) recorded in the fortnight
preceding pellet collection, i.e. the timeframe during which pellets were produced. We used
68
both proportion of prey items (%N), which represents hunting acts, and proportion of prey
biomass (%B), representing an index of energy intake by the owls. This latter index may be
relevant in an energetic context, especially during winter, an energetically stressful period
for raptors (e.g. Overskaug et al. 1997, Newton 1998). Estimates of prey
biomasses were derived from the literature (D i P alma & Massa 1981, Galeotti
& Canova 1994) and from specimens collected in the study area. Birds were assigned
a mean mass of 20 g each (Galeotti & Canova 1994, P irovano et al. 2000a).
The body mass of predated brown rats was estimated by measuring mandible length and
using the regression equation given in D i P alma & Massa (1981). In biomass
calculations, rats were thus assigned the mean monthly weight.
The proportions of prey categories (both as %N and %B) were calculated for each
fortnight. The number of prey items for each fortnight varied between 29 and 580 (median
= 198 prey items, see Appendix): overall, 25 fortnights were included in the analyses. All
variables were normally distributed (Kolmogorov-Smirnov test, P > 0.18), so that no data
transformation was performed before variables being used in parametric statistical
analyses. For each fortnight we calculated diet breadth using the L e v ins’ (1968) index
of niche breadth (NB = 1 / Σpi2, where piis the proportion of the ith prey category); compared
to other diversity indices, this index gives more weight to dominant diet components (which
can be hypothesized to have more biological relevance), and its value is less sensitive to
stochastic variation in the occurrence of the least abundant prey categories. Diet breadth was
calculated on all prey categories, including those not listed above.
In order to separate seasonal variation in diet composition from the effect of weather
(temperature and rainfall), we first built maximal models in which diet parameters were
expressed as a function of year of pellet collection and date of sampling. Given that
a preliminary look at the variation of prey categories in relation to date of sampling
revealed mostly non-linear patterns, we also included in the maximal models the quadratic
term of date, together with all the year by date interactions. Models were built using Type
I sum of squares (hierarchical sum of squares), in which each term is corrected only for
those preceding it in the model definition (N o r u s i s 1988). Thus, the factor year was
always entered first in the model, followed by date and its quadratic term, whereas
interactions (year x date, year x date2) were entered after main effects. Before testing the
additive effects of temperature and rainfall, non-significant higher order terms and
interactions were dropped from the model, and the model was run again with significant
terms included only (this is termed Step 1 hereafter). Note that if higher order terms were
significant, then the corresponding lower order terms were also left in the model, even if
they did not reach statistical significance. Then we tested for the effects of temperature
and rainfall, which were entered separately in the models (Step 2a and Step 2b,
respectively). If the initial model (before Step 1) was non significant, then the effects of
weather variables on diet parameters were tested alone in an analysis of covariance
(ANCOVA).
Owing to the strict non-linear covariation between temperature and date of sampling
(temperature as a function of date: R2= 0.73, 2nd order polymomial function, P < 0.0001),
with temperature being lower in mid-winter months, effects of seasonality per se and
temperature on diet could be difficult to disentangle. Hence, the same models as above
(Step 2a) were also run entering temperature before the effects of date of sampling: to
evaluate which model performed best, we compared the change in R2(∆R2) between
69
model types (i.e. those with temperature entered after date vs. those with temperature
entered before date) after entering the first term (note that seasonal terms, e.g. date and
date2, are considered altogether in calculating the change in R2). Means are shown together
with their standard deviations.
Results
In total, 5509 prey items were examined over the two years (see Appendix). The 2-wk
proportions of prey items and prey biomasses were strictly correlated (mean rs= 0.93; range
0.85–0.96; N = 5 prey categories; all P < 0.0001). Three out of five prey categories showed
a clear seasonal variation in prevalence in the diet (Table 1, Fig. 1), both in terms of number
and biomass (i.e. wood mouse, brown rat and harvest mouse). Birds showed a seasonal
variation in prevalence only when considering %N, whereas no variables were significant
in the initial model for %B (Table 1). Savi’s pine vole was the only prey category showing
a significant between-year variation in prevalence (Table 1). In both years, the dietary
response over the winter season was non-linear, the prevalence of rats and birds decreasing
around mid-winter being replaced by wood mice and birds. As mid-winter was the period
with the lowest temperatures, the prevalence of rats, harvest mice and birds trend could also
be modelled as a function of mean temperature (see significant effects at Step 2a in Table 1).
70
Table 1. Summary of ANCOVA models (type I sum of squares) describing the relationships between long-eared
owl diet parameters (diet composition, expressed both as % number (%N) and % biomass (%B)), seasonal
variation and weather variables during two consecutive study winters (October to April), 1996–1998. See Material
and Methods for a detailed description of variables and model definition.
Model steps
Diet parameters Step 1§(year, season) Step 2a^ (temperature) Step 2b^ (rainfall)
Variables R2(%) P ∆R2(%) P ∆R2(%)
Apodemus sylvaticus %N date, date270.7 0.25 - 0.14 - 0.68 -
%B date, date261.9 0.62 - 0.17 - 0.72 -
Rattus norvegicus %N date, date266.8 0.39 - 0.00* 66.8 - 14.5 0.10 -
%B date, date248.7 0.49 - 0.15 - 0.24 -
Microtus savii %N year 43.9 0.22 - 0.19 - 0.88 -
%B year, date 38.1 0.07 - 0.17 - 0.47 -
Micromys minutus %N date247.5 0.07 - 0.00* 47.5 - 51.3 0.39 -
%B date245.3 0.15 - 0.00* 45.3 - 41.8 0.32 -
Birds %N date, date259.6 0.29 - 0.02* 59.6 - 11.0 0.76 -
%B - - 0.49 - 0.10 -
§Only significant terms are shown; associate model R2is shown when overall models significant at P < 0.05.
^ P-values for additional effects of temperature and rainfall are shown; for temperature: left value: temperature
entered after other variables (those selected with Step 1); right value: temperature entered before other
variables. Significant P-values (P< 0.05) are marked with an asterisk (*). Corresponding changes in % R2(∆R2)
for the two model types (after entering the first term; seasonal terms are considered altogether in calculating the
change in R2, see Materials and Methods for details) are also shown when P-values for temperature were
significant at P < 0.05.
However, the only prey category that showed a consistent covariation in frequency with
mean temperature upon the general effect of seasonality was the harvest mouse, which
increased in proportion when temperature was lower, whereas seasonality seems to be far
more important for other prey categories (Table 1). No correlation was found between the
amount of rainfall and the percentages of any prey category (all P > 0.10, Table 1).
Diet breadth calculated on %N was not correlated with the same index calculated on
%B (rs= -0.11; N = 25; P = 0.59), suggesting that different factors could be affecting this
index at these two levels of analysis (%N or %B). The maximal model was non-significant
for %N diet breadth (F5,19 = 2.16, P = 0.10), whereas date (P = 0.004) and date2(P = 0.04)
were significant predictors of %B diet breadth after running Step 1, with breadth mostly
increasing from winter to spring (model: F2,22 = 7.59, P = 0.003). %N diet breadth
increased with decreasing temperature (F1,23 = 9.29, P = 0.006) (Fig. 2a) and with
increasing rainfall (F1,23 = 5.62, P = 0.026) (Fig. 2b), but there were no correlations
between weather variables and %B diet breadth (both entering weather variables after and
before date terms, all P > 0.05). The breadth indexes were not dependent on the total
number of prey items included in each fortnight sample (%N: rs= 0.30; N = 25; P = 0.14;
%B: rs= 0.08; N = 25; P = 0.70). Mean diet breadth calculated on %B was significantly
lower than that calculated on %N (2.12 ±0.50 vs. 3.53 ±0.59, respectively; paired t-test:
t24 = 8.57, P < 0.0001): this is because of the strong dominance of rats in biomass (mean
fortnight proportion of rat biomass in diet is 64.9 ±12.9 %, range 36.8–88.5 %, see Fig. 2,
Appendix). In fact, %B diet breadth was dependent on %B of rats in diet (rs= -0.98;
N = 25; P < 0.0001), emphasizing that the brown rat is by far the most important prey and
that the niche breadth in this apparently specialist predator is basically a function of the
prevalence in the diet of the most important single prey type. However, in this particular
situation, the Levins’ index calculated on %B, being heavily influenced by dominant prey
categories, may not be able to provide an adequate representation of the seasonal
variations in diet breadth.
Discussion
Owing to the proportional nature of dietary data, with one prey type increasing in
prevalence as others decrease, it may be difficult to identify the reasons behind changes in
diet composition (statistically as well as biologically). Indeed, from our investigations, it is
clear that the brown rat is the primary prey type in this wintering owl population, in which
birds behave as rat specialists (see also P irovano et al. 2000a). The prevalence of rats
reaches up to 80% of prey biomass and is around 20% in number of prey items. As
a consequence, the prevalence of other numerically important mammalian prey types
(wood mouse and Savi’s pine vole) were negatively correlated to the percentage of brown
rat and were thus seemingly alternative prey to the brown rat, given the dominance of rats
by biomass (P irovano et al. 2000a).
Overall, the most important prey categories showed clear within-season variations in
prevalence throughout the winter. Weather variables had no or limited effects on the
variation of individual prey categories: the only prey whose prevalence in diet was
consistently affected by temperature was the harvest mouse, which, owing to its small
size, makes a low contribution to the overall diet, and can thus be considered to have a low
biological relevance (see Appendix). The increased prevalence of harvest mice with cold
71
weather may reflect a higher availability of this prey with lower temperatures, when
a fraction of the population may be forced to leave cultivated fields by habitat
deterioration, thus showing a more predictable occurrence in network habitats largely
preferred by owls for hunting (Galeotti et al. 1997).
In general, the relationships between diet composition and weather conditions may
have different origins: either (1) predation by the owls is influenced by weather
conditions, given a uniform availability of prey species, or (2) activity/availability of prey
72
Fig. 1. Seasonal variation in diet composition of long-eared owls, measured at 2-wk intervals. For clarity of
presentation, only fitted lines for prey categories that showed significant seasonal variation in frequency are shown,
according to patterns detected in Table 1; a) frequency of prey items (%N); b) frequency of prey biomass (%B).
species depends on weather variations and the owls eat what is available. Of course, these
possible processes are not mutually exclusive, and what is actually found in the owls’ diet
may derive form their combination: a thorough investigation of the factors influencing diet
should link dietary data to small mammal and owl’s activity pattern and habitat use. Future
efforts should thus include the quantification of the above-mentioned variables, and
perhaps simultaneous trapping of small mammals in different habitats and quantification
of owls habitat use (by e.g. radio-tracking), while taking into account weather conditions,
may provide adequate answers.
73
Fig. 2. Correlation between an index of long-eared owl diet breadth (Levins’ index), calculated for each fortnight
on %N, and: a) 2-wk mean ambient temperature; b) total amount of fortnight rainfall (in mm); filled circles =
winter 1996–97; open circles = winter 1997–98.
With regards to our results, possible reasons for the observed generalized lack of effect
of weather variables on individual prey frequencies may include the relatively mild winter
climate and the interval at which pellet were collected, that may have been too long and
may have obscured the effect of weather on diet. Previous studies of resident raptors
wintering at high latitudes have shown food availability, related to climatic factors, to be
of major importance in determining the survival probability throughout winter (Newton
1998, Sunde 2002), but weather conditions are likely to have a limited importance in
our study area on both owl activity and food abundance: hence, it would be interesting to
investigate longer time series including harsh winter seasons, which may exacerbate the
effect of weather on diet composition.
Diet breadth (calculated on %N) was positively related to rainfall and negatively to
temperature. This may simply reflect a decrease in diet breadth when many young rats are
available, i.e. when temperatures are high (P erry 1946, P irovano et al. 2000a):
however, this would imply a negative correlation between %N diet diversity and
abundance of rats (both %N and %B), but this is not the case (Spearman Rank: P = 0.60
and P = 0.30, respectively). Hence, either overall average prey activity may be higher at
lower temperatures and elevated soil moisture (M aguire 1999), or owls may become
more euryphagous under inclement weather conditions, which seems a likely explanation
(Canova 1989). In fact, lower prey selectivity under relatively unfavourable weather
conditions, which may negatively affect the energy budget of the owls or the abundance of
food resources, is to be a generally expected ecological response to climatic variability
(e.g. MacArthur & Pianka 1966). When analysing diet diversity from an
energetical point of view (%B diet breadth), given that rats dominate the diet, diversity
increased when the prevalence of rats decreased during mid-winter months (probably in
response to a lower availability of young rats), with owls shifting to other prey types
(wood mouse, Savi’s pine vole). Thus, diet diversity seems to be influenced both by
climatic factors and by the abundance of the preferred prey, depending on the measure
used (%N or %B).
To conclude, the diet composition of the long-eared owl in our study area appeared to
be independent of rainfall and slightly influenced by temperature, while a measure of diet
variability increased with decreasing temperature and with an increasing amount of
rainfall. Taken together, these results suggest that further studies investigating the diet of
owls should consider seasonal variation and weather conditions as potentially influential
variables, and emphasize that the winter diet composition of the long-eared owl in
Southern Europe show marked within–season fluctuations.
Acknowledgements
The content of pellets was determined by S. Brambilla. We thank all the people who helped with fieldwork
and G. Bogliani, P. Galeotti, C. Maguire, C. Marti and an anonymous reviewer for suggestions
on a previous draft of the manuscript. We are indebted with P. Sunde for his insightful comments and for
thorough discussion on data analysis. O. Janni kindly checked the English. We are also grateful to the Editor
and Dr G. Copp for comments on the final revision.
74
75
Appendix. Diet parameters (for both N and B) of long-eared owls in the suburbs of Milan, northern Italy, from October to April, according to year (Y) (1 = winter 1996/97;
2 = winter 1997/98) and date (D) of collection (1 = 1-15 October; 14 = 16-30 April). Diet breadth, number of prey items considered (N items) and weather variables (Temp =
temperature; rain = rainfall) are also shown (see Material and Methods for details).
Y D Apodemus sylvaticus Rattus norvegicus Microtus savii Micromys minutus Birds Diet N Temp Rain
breadth§items (°C) (mm)
%N %B %N %B %N %B %N %B %N %B N B
1 3 29.41 9.46 42.53 81.74 10.41 3.04 3.17 0.26 9.50 2.89 3.33 1.46 221 11.10 55.0
1 4 22.22 11.20 18.18 58.24 33.33 15.24 0.00 0.00 19.19 9.14 4.25 2.57 99 6.21 90.6
1 5 21.89 9.66 22.26 68.00 33.21 13.30 2.64 0.30 13.58 5.66 4.33 2.02 265 6.38 68.0
1 6 34.48 12.68 31.03 78.98 10.34 3.45 6.90 0.65 6.90 2.40 4.14 1.56 29 2.59 57.8
1 7 44.22 24.27 13.27 58.07 21.09 10.50 11.22 1.57 6.12 3.17 3.62 2.43 294 2.69 79.2
1 8 53.85 33.51 9.89 47.44 22.53 12.71 5.49 0.87 2.75 1.61 2.78 2.74 182 5.12 20.2
1 9 43.28 23.30 13.79 59.13 23.45 11.45 8.28 1.14 4.48 2.28 3.61 2.33 580 5.91 0.0
1 10 43.17 20.24 21.03 66.03 23.25 9.89 3.32 0.40 4.43 1.96 3.37 2.01 271 8.81 10.8
1 11 52.71 28.09 17.44 58.11 18.22 8.81 3.10 0.42 3.49 1.76 2.84 2.29 258 12.67 0.0
1 12 34.43 13.95 28.96 73.57 19.13 7.03 2.73 0.28 12.02 4.60 3.93 1.76 183 12.87 0.0
1 13 50.88 25.24 21.05 62.42 17.54 7.89 1.75 0.22 6.14 2.88 2.86 2.11 114 13.94 0.0
1 14 27.91 15.97 16.28 55.69 37.21 19.31 0.00 0.00 6.98 3.77 3.62 2.44 43 12.54 16.6
2 1 9.38 3.12 42.19 82.04 1.56 0.47 0.00 0.00 34.38 10.82 3.20 1.45 64 18.41 10.2
2 2 15.24 4.21 55.49 88.50 6.71 1.68 2.44 0.17 10.98 2.86 2.85 1.27 164 10.75 10.4
2 3 31.31 11.22 38.89 80.39 3.54 1.15 3.03 0.28 15.15 5.13 3.62 1.52 198 9.81 114.2
2 4 25.82 8.34 39.56 83.16 10.44 3.06 3.85 0.32 11.54 3.52 3.97 1.43 182 6.69 5.0
2 5 35.20 13.96 26.20 75.10 10.80 3.88 9.60 0.97 9.20 3.44 4.42 1.70 500 5.74 51.4
2 6 43.23 27.62 10.42 51.16 14.84 8.60 9.11 1.48 12.24 7.38 4.04 2.85 384 3.36 127.4
2 7 44.31 24.21 16.08 58.76 14.90 7.38 5.49 0.76 7.84 4.05 3.84 2.42 255 3.82 42.2
2 8 35.40 17.69 17.31 64.79 18.60 8.43 8.53 1.09 10.34 4.88 4.69 2.17 387 4.10 16.8
2 9 48.73 29.38 12.39 51.42 21.97 12.01 4.79 0.74 9.01 5.13 3.21 2.72 355 6.47 0.0
2 10 49.38 22.67 20.33 66.85 14.11 5.87 3.32 0.39 5.81 2.52 3.20 1.99 241 10.98 39.0
2 11 54.55 41.10 6.82 36.79 19.32 13.20 6.82 1.31 7.95 5.66 2.84 3.07 88 9.96 1.6
2 12 58.41 28.35 22.12 64.01 13.27 5.84 3.54 0.44 1.77 0.81 2.44 2.03 113 10.34 9.6
2 13 48.72 26.92 17.95 53.48 12.82 6.42 5.13 0.72 7.69 4.01 3.37 2.72 39 11.12 71.4
§According to the Levins’index of niche breadth (NB = 1 / ∑pi2) on all prey categories determined to species level (including birds, that were considered as a single category).
LITERATURE
BARBIERI F., BOGLIANI G. & FASOLA M. 1975: Aspetti della predazione di un barbagianni (Tyto a. alba
Scop.). In: Atti V Simposio Nazionale sulla Conservazione della Natura, Vol. I. Istituto di Zoologia,
Universitàdi Bari: 293–302.
CANOVA L. 1989: Influence of snow cover on prey selection by Long-eared Owl Asio otus. Ethol. Ecol. Evol. 1:
367–372.
DI PALMA M. G. & MASSA B. 1981: Contributo metodologico per lo studio dell’alimentazione dei rapaci. In:
Farina A. (ed.), Atti I° Convegno Italiano di Ornitologia. Mus. St. Nat. Lunigiana, Aulla: 69–76.
GALEOTTI P. & CANOVA L. 1994: Winter diet of Long-eared Owl (Asio otus) in the Po Plain (Northern Italy).
J. Raptor Res. 28: 265–268.
GALEOTTI P., TAVECCHIA G. & BONETTI A. 1997: Home-range and habitat use of long-eared owls in open
farmland (Po Plain, northern Italy), in relation to prey availability. J. Wildl. Res. 2: 137–145.
GETZ L. L. 1961: Responses of small mammals to live traps and weather conditions. Amer. Midl. Nat. 66:
160–170.
HENRIOUX F. 1999: Écologie d’une population de Hiboux moyen-ducs Asio otus en zone d’agriculture
intensive. Ph.D. dissertation, Université de Neuchâtel, Switzerland, 119 pp.
LEVINS R. 1968: Evolution in changing environments. Princeton University Press, Princeton, 132 pp.
MACARTHUR R. H. & PIANKA E. 1966: On optimal use of a patchy environment. Amer. Nat. 100: 603–609.
MAGUIRE C. C. 1999: Rainfall, ambient temperature, and Clethrionomys californicus capture frequency.
Mammal Rev. 29: 135–142.
MARTI C. D. 1976: A review of prey selection by the Long-eared Owl. Condor 78: 331–336.
MICHELAT D. & GIRAUDOUX P. 1992: Activité nocturne et stratégie de recherche de nourriture de la chouette
effraie Tyto alba àpartir du site de nidification. Alauda 60: 3–8.
NEWTON I. 1998: Population Limitation in Birds. Academic Press, London, 597 pp.
NILSSON I. N. 1981: Seasonal changes in food of the Long-eared Owl in southern Sweden. Ornis Scand. 12:
216–223.
NORUSIS M. J. 1988: SPSS Guide to Data Analysis for SPSS/PC+. SPSS Inc., Chicago, 416 pp.
OLSEN P. & OLSEN J. 1992: Does rain hamper hunting by breeding raptors? Emu 92: 184–187.
OVERSKAUG K., SUNDE P. & KRISTIANSEN E. 1997: Subcutaneous fat accumulation in Norwegian owls and
raptors. Ornis Fenn. 74: 29–37.
PERRY J. S. 1946: The reproduction of the wild brown rat (Rattus norvegicus Erxleben). Proc. Zool. Soc.,
London, 115: 19–46.
PIROVANO A., RUBOLINI D., DE MICHELIS S. & FERRARI N. 1997: Primi dati sull’ecologia di un roost di
Gufo comune Asio otus in ambiente urbano. Avocetta 21: 89.
PIROVANO A., RUBOLINI D., BRAMBILLA S. & FERRARI N. 2000a: Winter diet of urban roosting Long-
eared Owls Asio otus in Northern Italy: the importance of the Brown Rat Rattus norvegicus. Bird Study 47:
242–244.
PIROVANO A., RUBOLINI D. & DE MICHELIS S. 2000b: Winter roost occupancy and behaviour at evening
departure of urban long-eared owls. Ital. J. Zool. 67: 63–66.
SIDOROWICZ J. 1960: Influence of the weather on capture of Micromammalia. I. Rodents (Rodentia). Acta
Theriol. 4: 139–158.
SUNDE P. 2002: Starvation mortality and body condition of Goshawks Accipiter gentilis along a latitudinal
gradient in Norway. Ibis 144: 301–310.
WIJNANDTS H. 1984: Ecological energetics of the Long-eared Owl (Asio otus). Ardea 72: 1–92.
YALDEN D. W. & MORRIS P. A. 1990: The Analysis of Owl Pellets. Mammal Society Occasional Publication
No. 13, London, 24 pp.
76
