Science - USA (2022-02-18)

Interactions with natural and anthropogenic
environments
Coupled with fine-scale environmental mon-
itoring, high-throughput tracking systems re-
veal how animals respond to environmental
stimuli (Fig. 4 and movies S3 to S5), providing
critical information for developing effective

management and restoration actions ( 13 , 14 ).

For example, high-resolution GPS data (dt= 1 s)

combined with triaxial accelerometry and

atmospheric modeling were necessary to re-

veal differential responses of adult and juvenile

griffon vultures (Gyps fulvus) to challenging

soaring conditions ( 38 ) (Fig. 4A and movie S3).

Further, whole-lake acoustic trilateration (dt=

9 s) revealed interaction between a non-native

wels catfish (Silurus glanis) and physical

features (e.g., water temperature) of a novel

environment ( 27 ).

High-throughput tracking data, coupled

with mapping of relevant human activities,

Nathanet al.,Science 375 , eabg1780 (2022) 18 February 2022 4 of 12

Non woodland

Woodland

Post-mortem (fox) movement

Live (pheasant) movement

Death

25-Sep-2018 21:34

27-Sep-2018 06:26

25-Sep-2018 21:35

-3.902 -3.898

(^27)
(^7).
0
5
(^37)
(^7).
(^05)
Carcass found
03-Oct-2018 15:48
A
Speed (m*s
-1)
01 Sep 30 Sep
5
(^0).
0
(^1).
0
0
Slow Fast
B
0 50 m
25-Sep-2018 12:30
Survival
Average daily swimming distance (m)
1000 2000 3000
Captured
Uncaptured
Captured by angling Never captured by angling
βμ = -0.368
Time
15 Sep
50.769
50.771
50.769 50.772
-3.906 -3.901
Early Late
-3.906 -3.901
0 100 m
-3.901 -3.901
Good spatial
cognition performance
Poor spatial
cognition performance
1 5 1 0 02 3 0 04 5 0 06
-0.5
-0.4
-0.3
Sampling interval (min)
Linear mean standardized
selection gradient (
β
μ
)
52.992 13.578 13.588
52.996
0 100 m

Fig. 2. Inference on patterns of variation in movement, behavior, and
fitness among individuals and their potential drivers.(A) ATLAS-tracked
(dt= 4 s) young pheasants (P. colchicus) that performed better in spatial
cognitive tasks in captivity made slower transitory movements during early
stages of exploration in the wild, but their speed increased with experience
in the environment; poor cognitive performers moved faster during early
exploration but did not differ in their speed later on ( 32 ) (top plot). This
general trend is illustrated for two representative ATLAS-tracked individuals
(middle plots). Histograms show the number of fast steps (>1 m/s). The
bottom map shows the track of a pheasant (blue lines) that was killed and
carried away (with the ATLAS tag intact) by an untagged fox (Vulpes vulpes)
(black lines). ATLAS informed the exact timing and location of such mortality
events, whereas in situ observations (skull and crossbones, magnifying glass)
would place the mortality location 400 m away with an 8-day uncertainty about
its timing in this example. (B) More active northern pike (E. lucius) tracked
in the wild using acoustic trilateration (dt= 9 s) were more likely to be
captured by angling (purple) (top plot), suggesting that angling pressure
results in shyer, less active pike populations (blue) ( 30 ). Variation in
activity between captured and noncaptured pike is illustrated in the map
by six representative tracks (marked by asterisks in the top plot), with dotted
lines representing data gaps (dt> 60 s). The strength of harvest selection
on fish behavior, represented by the mean-standardized linear selection
gradient (bm), is rapidly overestimated (more negative values) as temporal
resolution decreases (longer sampling intervals) (bottom plot).
RESEARCH | REVIEW