\documentclass[11pt]{article}

\usepackage[margin=1in]{geometry}
\usepackage{amsmath}
\usepackage{amssymb}
\usepackage{graphicx}
\usepackage{booktabs}
\usepackage{tikz}
\usepackage{pgfplots}
\pgfplotsset{compat=1.18}
\usepgfplotslibrary{errorbars}
\usepackage[colorlinks=true, linkcolor=black, citecolor=black, urlcolor=blue]{hyperref}
\usepackage{textcomp}
\newcommand{\permil}{\textperthousand}

% --- skin palette: night indigo, brass sonar, robin rust -------------------
\definecolor{night}{HTML}{34467A}
\definecolor{nightdeep}{HTML}{202F57}
\definecolor{sonar}{HTML}{A3771C}
\definecolor{rust}{HTML}{A83F2C}
\definecolor{evid}{HTML}{1C8F6A}

\title{Twenty-Three Minutes of Chewing, One Kilometre Up}
\author{The Science Journaling Club}
\date{Field Note --- Paper Analysis --- Animal Behaviour}

\begin{document}
\maketitle

\begin{abstract}
Carlos Ib\'a\~nez found the first clue in 2001: feathers, too many to be
nesting material, packed into the guano beneath a greater noctule bat roost in
southern Spain \cite{ibanez2001}. In 2007 his group ran the blood chemistry and
got an isotope signature that, every autumn, sat inside the zone reserved for
animals eating almost nothing but birds \cite{popalisseanu2007} --- a gap of
$3.21\permil$ in carbon and $3.15\permil$ in nitrogen from the bats' ordinary
insect diet, our own subtraction of the paper's reported means. Neither result
could say \emph{how} a bird ends up eaten by something that echolocates.
Caught alive, a kilometre up, by an animal the size of a fist? Or found already
dead? Isotopes cannot tell a kill from a carcass; the molecules read the same
either way. For eighteen more years nobody could resolve it, because nobody
could follow a bat far enough into the dark to watch. In October 2025 a team
led by Laura Stidsholt did, by making the bat carry the witness. Sound-and-
movement tags built at Aarhus University, riding on \textit{Nyctalus
lasiopterus} over Spain's Do\~nana reserve, recorded two hunts start to finish:
a 30-second chase that failed, and a chase lasting nearly three minutes ---
more than a kilometre above the ground, in total darkness --- that ended with
21 recorded distress calls from the bird, a robin, followed by 23 minutes of
chewing sounds as the bat kept flying and ate it in the air
\cite{stidsholt2025}. This Field Note follows the case from the guano pile to
the recording, and asks the question the tag actually answers: what does a
hunting technique built for insects a few centimetres long look like when it
is redirected at something built to survive a hawk?
\end{abstract}

%============================================================================
\section{A Guano Pile That Started an Argument}

Do\~nana sits at the southwestern corner of Spain, marshland and pine and
cork-oak scrub against the mouth of the Guadalquivir, and every autumn the sky
above it fills with something that never touches the ground there: millions of
small songbirds flying south at night on their way to Africa. Below them roosts
one of the stranger residents of the same landscape. \textit{Nyctalus
lasiopterus}, the greater noctule, is the largest bat in Europe, with a
wingspan that can reach 46~cm \cite{phys2025}.

Greater noctule colonies at Do\~nana are themselves an unusual social system to
be attached to this story: the same research group has mapped their
roost-switching as a highly structured fission--fusion network, individuals
splitting into and rejoining different tree-cavity groups on a pattern closer
to what is normally described for primates than for bats
\cite{popalisseanu2008,fortuna2009}. It is the same population, and largely the
same research lineage, that later supplied both the isotope case and the
recording that closed it.

In 2001 Carlos Ib\'a\~nez and colleagues published something odd about it.
Faecal pellets collected below greater noctule roosts contained bird feathers
--- not a trace, but a recurring, seasonal presence, timed to exactly the weeks
when migratory songbirds pass overhead at night \cite{ibanez2001}. A bat roost
accumulating bird feathers is not, on its own, proof of anything specific:
birds die of many causes during a migration that can run thousands of
kilometres, and a bat that merely scavenges an already-dead bird would leave
feathers in its droppings too. The 2001 paper established that greater
noctules were, somehow, involved with dead birds. It could not establish
whether the bat was the reason a given bird was dead. That distinction stayed
open for twenty-four years.

%============================================================================
\section{What the Isotopes Could Prove, and What They Couldn't}

The next piece of evidence, in 2007, came from chemistry. Popa-Lisseanu,
Ib\'a\~nez and colleagues measured stable carbon and nitrogen isotope ratios,
$\delta^{13}$C and $\delta^{15}$N, in greater noctule tissue and compared them
against reference values for the two candidate food sources: flying
invertebrates, and the migrating passerines themselves \cite{popalisseanu2007}.
Passerine tissue: $\delta^{13}$C $= -22.43 \pm 0.32\permil$, $\delta^{15}$N
$= 9.60 \pm 0.26\permil$. Invertebrate tissue: $\delta^{13}$C $= -25.64 \pm
0.23\permil$, $\delta^{15}$N $= 6.45 \pm 0.39\permil$ \cite{popalisseanu2007}.
Subtracting one from the other --- our own arithmetic, not the paper's --- the
two diets sit $3.21\permil$ apart in carbon and $3.15\permil$ apart in
nitrogen. Autumn greater noctule blood fell in the lower part of the
theoretical box representing a nearly pure-bird diet, and feathers made up
roughly half of faecal pellet content in that season \cite{popalisseanu2007}.

\begin{figure}[htbp]
\centering
\begin{tikzpicture}
\begin{axis}[
  width=12cm, height=7cm,
  xmin=-27, xmax=-21, ymin=5, ymax=11,
  xlabel={$\delta^{13}$C (\textperthousand)},
  ylabel={$\delta^{15}$N (\textperthousand)},
  ymajorgrids=true, grid style={dashed, gray!30}
]
\addplot[evid, only marks, mark=*, mark size=3.2pt,
  error bars/.cd, x dir=both, x explicit, y dir=both, y explicit]
  coordinates { (-25.64,6.45) +- (0.23,0.39) };
\addlegendentry{invertebrates}
\addplot[rust, only marks, mark=*, mark size=3.2pt,
  error bars/.cd, x dir=both, x explicit, y dir=both, y explicit]
  coordinates { (-22.43,9.60) +- (0.32,0.26) };
\addlegendentry{passerines}
\addplot[gray, dashed, mark=none] coordinates {(-25.64,6.45) (-22.43,9.60)};
\end{axis}
\end{tikzpicture}
\caption{The isotope evidence from 2007, plotted from the paper's reported
means and standard deviations \cite{popalisseanu2007}. The two prey types
separate cleanly, enough that the authors could place autumn bat tissue inside
the theoretical zone for a nearly pure-bird diet. What the plot cannot show is
a bat's own sample landing on it, because isotopes carry no information about
whether the bird supplying that signature was alive when the bat found it.}
\label{fig:isotope}
\end{figure}

The 2007 paper is careful on exactly this limit: the isotopes show that
greater noctules were, seasonally, eating birds almost exclusively. They
cannot show whether those birds were caught in flight or found already dead.
A molecule of bird protein carries no memory of how the bird died.

%============================================================================
\section{The Backpack That Finally Rode Along}

What closes a gap like that is not a cleverer assay; it is a witness. For an
animal flying fast, in the dark, more than a kilometre up, no human observer
and no camera can be that witness. The only candidate observer that can go
where the bat goes is the bat. A biologging tag is an instrument small and
light enough to ride on a wild animal without materially changing what it
does, recording what the animal itself experiences. Stidsholt's team fitted
greater noctules at Do\~nana with tags built at Aarhus University that
recorded movement, acceleration, altitude and sound --- including the bat's own
echolocation calls --- on sub-second timescales \cite{stidsholt2025,phys2025}.
The tag does not need to guess what the bat is doing. It hears the bat
deciding.

%============================================================================
\section{Two Chases, Recorded Whole}

Across the tagging effort, the team recovered two full predation attempts on
tape: a chase that failed, and a chase that ended in a kill
\cite{stidsholt2025,phys2025}. Two events is a thin sample, but each is not a
data point in the usual sense --- each is a continuous acoustic and inertial
record of an entire hunt, second by second, of a kind nobody had obtained for
this behaviour before.

\begin{figure}[htbp]
\centering
\begin{tikzpicture}
\begin{axis}[
  width=11cm, height=6.5cm,
  ybar, bar width=28pt,
  ymin=0, ymax=200,
  ylabel={duration (s)},
  symbolic x coords={Chase A (no capture),Chase B (kill)},
  xtick=data,
  ymajorgrids=true, grid style={dashed, gray!30},
  nodes near coords
]
\addplot[fill=night] coordinates {(Chase A (no capture),30) (Chase B (kill),180)};
\end{axis}
\end{tikzpicture}
\caption{The two recorded pursuits, reported at 30 seconds and ``nearly three
minutes'' \cite{phys2025}; we take the second as $\approx180$~s and flag it as
approximate. Both were flown more than a kilometre up in total darkness
\cite{phys2025}. The successful chase ran roughly six times longer than the
failed one --- our own division of the two durations.}
\label{fig:chases}
\end{figure}

%============================================================================
\section{The Same Trick Bats Use on Moths}

To read what the tag's audio track means, it helps to know what an ordinary
echolocating bat hunt sounds like, because the greater noctule did not invent
a new strategy for this occasion. Since Griffin, Webster and Michael's classic
1960 description, bat biologists have recognised a standard three-part attack
sequence: a search phase, with calls spaced well apart while the bat scans open
air; an approach phase, as calls quicken once a target is detected and
tracked; and a terminal buzz, an extremely rapid burst of very short calls
fired in the last instants before contact \cite{griffin1960,surlykke2008}.

What makes the 2025 recording remarkable is not that a buzz occurred, but what
it was aimed at. This acoustic strategy was shaped, over a long evolutionary
history, for prey a few centimetres long and a fraction of a gram, resolved at
close range in encounters that insect-hunting studies describe as
characteristically brief. The tag caught the same signature --- ``a rapid
burst of short echolocation calls'' \cite{phys2025} --- turned on something
that outweighs most insects a bat has ever eaten by orders of magnitude,
sustained across a pursuit lasting minutes rather than an instant. The toolkit
did not change. The target did.

%============================================================================
\section{Twenty-One Distress Calls}

The successful chase ran more than a kilometre above the ground, in total
darkness, for close to three minutes \cite{stidsholt2025,phys2025}. Near the
end of it, the tag's microphone caught something no isotope assay or guano
sample ever could: the bird itself, alive and audibly distressed. Twenty-one
distress calls from the robin were recorded around the capture
\cite{phys2025}. A bird does not vocalise in alarm twenty-one times over the
course of being found already dead. This single fact does the most work in the
entire case: a recording of the prey's own voice, mid-attack, cannot be
explained by scavenging at all.

%============================================================================
\section{Twenty-Three Minutes of Chewing, Still Airborne}

After the capture, the bat did not land, and it did not stop recording. It
descended toward lower altitude and kept flying while it ate, and the tag's
microphone picked up twenty-three minutes of chewing sounds as it did
\cite{phys2025}. Add that to the roughly three-minute chase and the whole
documented event, start to finish, comes to about 1{,}560 seconds --- 26
minutes --- of continuous flight: our own addition of the two reported
durations, not a number either paper states outright.

\begin{figure}[htbp]
\centering
\begin{tikzpicture}[line width=1pt]
  \draw (0,0) -- (12,0);
  \draw[night, line width=10pt] (0,0) -- (0.42,0);
  \draw[rust, line width=1.6pt] (0.42,-0.35) -- (0.42,0.35);
  \draw[sonar, line width=10pt, opacity=0.55] (0.42,0) -- (12,0);
  \node[font=\scriptsize, above] at (0.21,0.5) {chase, 180 s};
  \node[font=\scriptsize, above] at (0.42,0.55) {capture: 21 distress calls};
  \node[font=\scriptsize, above] at (6.2,0.5) {feeding, still flying --- 23 min of chewing sounds};
  \node[font=\scriptsize, below] at (0,-0.5) {0 min};
  \node[font=\scriptsize, below] at (12,-0.5) {26 min};
  \node[font=\scriptsize, below] at (6,-1.1) {total documented: 180 s + 1{,}380 s = 1{,}560 s $\approx$ 26 min (our addition)};
\end{tikzpicture}
\caption{The full arc of the successful hunt, reconstructed from the two
durations and the capture-moment detail the paper reports
\cite{stidsholt2025,phys2025}. Nearly all of the recorded event is the animal
eating on the wing, not chasing.}
\label{fig:timeline}
\end{figure}

Handling prey that size in flight is its own engineering problem. According to
the account of the recording, the bat bites off the bird's wings, most
plausibly to cut weight and drag, and manages the remainder using its
uropatagium --- the membrane between its hind legs and tail --- pulled forward
into a pouch to hold and eat the body while still airborne \cite{phys2025}.
That pouching motion is not a new invention: aerial-hawking bats routinely use
the same tail membrane to catch and manage insect prey mid-flight. What
changed here is the scale of what went into the pouch.

%============================================================================
\section{A Kilometre of Empty, Undefended Air}

Both recorded hunts took place more than a kilometre above the ground, in
total darkness \cite{stidsholt2025,phys2025}. That altitude is close to the
whole point of the story. Diurnal raptors hunt by sight and by day. Owls hunt
largely closer to the ground and to cover. Migratory songbirds fly at altitude
specifically at night, which removes them from most of the predators that
would otherwise be waiting. A greater noctule, hunting by echolocation in
total darkness a kilometre up, operates in a layer of night air that
essentially nothing else with wings is set up to hunt in at the same time.
Sonar does not care that there is no moon.

%============================================================================
\section{Why This Was Never Scavenging}

Return to the question the 2007 isotope paper could not close. Isotopes
describe diet; they cannot describe the manner of a prey animal's death,
because the chemistry of a bird's tissue does not remember whether the bird
was alive minutes before it entered a predator's stomach. The 2025 recording
instead supplies a continuous behavioural record of the entire event: a
directed pursuit lasting on the order of three minutes, tracked throughout by
the bat's own echolocation and inertial sensors, ending in a capture
accompanied by twenty-one distress calls from a bird that was, self-evidently,
alive and resisting right up to that point \cite{stidsholt2025,phys2025}. A
carcass picked off a branch does not require a three-minute pursuit. A carcass
does not scream. Put plainly: the isotopes told the field what the bats were
eating. The tag told it how.

%============================================================================
\section{Half Its Own Weight, Never Landing}

Asked about the size mismatch, lead author Laura Stidsholt offered a
comparison worth repeating in her own terms: ``A bird like that weighs about
half as much as the bat itself --- it would be like me catching and eating a
35-kilo animal while jogging'' \cite{phys2025}. That is the researcher's own
analogy, not a figure from the paper's methods, and we quote it as exactly
that. But the shape of the comparison is useful: most flying predators that
hunt on the wing take prey a small fraction of their own body mass. A prey
item at roughly half the predator's own mass, subdued and then eaten without
ever landing, is a different category of feat --- and it is the category the
tag shows the greater noctule attempting twice in the same dataset, once
without success and once, at length, with it.

%============================================================================
\section{What Changed, and What the Two Recordings Can't Tell Us}

Twenty-four years of work changed the type of evidence in hand: from a pile of
feathers, to a chemical signature, to a continuous audio-and-movement
recording of one predator killing and eating one prey animal, entirely in the
air. Carlos Ib\'a\~nez is a co-author on both the 2001 paper that opened the
case and the 2025 paper that closed it \cite{ibanez2001,stidsholt2025}. Two
recorded events is not a large sample. But the standard a single recording of
this kind has to clear is not statistical; it has to demonstrate that the
behaviour is physically possible and that it happened at least once under
conditions a scavenger could not produce. Twenty-one distress calls mid-
pursuit does that on its own.

What it cannot tell us: how often this happens across a greater noctule's
whole diet and season; whether the acoustic pattern of a bird-hunting buzz
differs, at the level of individual call timing, from an insect-hunting one;
or whether related bird-eating bats elsewhere --- a molecular diet study has
already found bird remains in a related noctule species in Japan
\cite{ibanez2021jp} --- hunt the same way. The tag answered the question it
was built to answer. It also, as tags tend to, opened several it wasn't.

\emph{A note on arithmetic.} Everything numerical in this Field Note that is
not a direct quantity from \cite{stidsholt2025}, \cite{ibanez2001},
\cite{popalisseanu2007} or the press account in \cite{phys2025} is flagged
inline as the club's own arithmetic: the $3.21\permil/3.15\permil$ isotope
gap, the $6\times$ chase-duration ratio, the 1{,}560~s / 26-minute total, and
the 24/18/6-year gaps between publications.

%============================================================================
\begin{thebibliography}{99}

\bibitem{stidsholt2025}
Stidsholt, L., Tena, E., Foskolos, I., Nogueras, J., de la Hera, I.,
S\'anchez-Navarro, S., Garc\'ia-Mudarra, J.~L. \& Ib\'a\~nez, C. (2025).
Greater noctule bats prey on and consume passerines in flight.
\textit{Science} 390, 178--181. doi:10.1126/science.adr2475

\bibitem{ibanez2001}
Ib\'a\~nez, C., Juste, J., Garc\'ia-Mudarra, J.~L. \& Agirre-Mendi, P.~T.
(2001). Bat predation on nocturnally migrating birds.
\textit{Proceedings of the National Academy of Sciences} 98, 9700--9702.
doi:10.1073/pnas.171140598

\bibitem{popalisseanu2007}
Popa-Lisseanu, A.~G., Delgado-Huertas, A., Forero, M.~G., Rodr\'iguez, A.,
Arlettaz, R. \& Ib\'a\~nez, C. (2007). Bats' Conquest of a Formidable Foraging
Niche: The Myriads of Nocturnally Migrating Songbirds. \textit{PLoS ONE} 2,
e205. doi:10.1371/journal.pone.0000205

\bibitem{griffin1960}
Griffin, D.~R., Webster, F.~A. \& Michael, C.~R. (1960). The echolocation of
flying insects by bats. \textit{Animal Behaviour} 8, 141--154.
doi:10.1016/0003-3472(60)90022-1

\bibitem{surlykke2008}
Surlykke, A. \& Kalko, E.~K.~V. (2008). Echolocating bats cry out loud to
detect their prey. \textit{PLoS ONE} 3, e2036.
doi:10.1371/journal.pone.0002036

\bibitem{popalisseanu2008}
Popa-Lisseanu, A.~G., Bontadina, F., Mora, O. \& Ib\'a\~nez, C. (2008).
Highly structured fission--fusion societies in an aerial-hawking, carnivorous
bat. \textit{Animal Behaviour} 75, 471--482.
doi:10.1016/j.anbehav.2007.05.011

\bibitem{fortuna2009}
Fortuna, M.~A., Popa-Lisseanu, A.~G., Ib\'a\~nez, C. \& Bascompte, J. (2009).
The roosting spatial network of a bird-predator bat. \textit{Ecology} 90,
934--944. doi:10.1890/08-0174.1

\bibitem{ibanez2021jp}
Ib\'a\~nez, C., Fukui, D., Popa-Lisseanu, A.~G., Pastor-Bevi\'a, D. \& Juste,
J. (2021). Molecular identification of bird species in the diet of the
bird-like noctule bat in Japan. \textit{Journal of Zoology} 313, 276--282.
doi:10.1111/jzo.12855

\bibitem{phys2025}
Phys.org (2025). How Europe's largest bat catches and eats birds mid-air.
News report on \cite{stidsholt2025}, including comment from L. Stidsholt.

\end{thebibliography}

\end{document}