Our fourth and final aim was to address the
synergistic effects of defaunation and climate
change by estimating how altered bird and
mammal assemblages influence the capacity
of plants to track local climate change through
seed dispersal. Dispersal is a crucial aspect of
biotic climate adaptation for plants ( 30 ), and
dispersal limitation, along with demographic
and competitive constraints, underlies lagged
responses of plants to climate change ( 31 ).
Because propagule pressure is a key factor in
range expansion ( 32 ), we sought to quantify
“climate-tracking dispersal,”or the ability of
seed dispersal to track climate change. We
conceptualized this problem by considering
how many seeds removed by frugivores from
a plant during a given year’s fruiting season
disperse to distances that exceed the average
distance that climate isoclines shift during a
year—that is, climate change velocity measured
in kilometers per year ( 11 ). For each grid cell,
we calculated the climate-tracking dispersal
index based on modeled dispersal estimates
involving a representative plant species with
median trait values, the animal species within
each grid cell, and its observed recent climate
velocity. The climate change velocities within
each pixel were calculated by averaging across
mean annual temperature and total annual
precipitation velocities observed for the period
1975 – 2013 ( 33 ).
We found that climate-tracking disper-
sal is especially limited in temperate regions
and areas with little topographic complexity
(Fig. 4A). We investigated how climate-tracking
dispersal function has been affected by past
defaunation and how it is threatened by the
current endangerment of birds and mammals
(Fig. 4B). Globally, we estimate that past
species losses have caused a 59.7% average
reduction in climate-tracking dispersal, and
the future loss of vulnerable and endangeredspecies from their current ranges would re-
sult in a further reduction of 15% globally.
Substantial spatial variation exists in the loss
of, and threat to, climate-tracking dispersal.
For example, dispersal—even by small birds—
can track the relatively low climate velocities
in the tropical Andes, and this region has lost
little of its climate-tracking dispersal function.
By contrast, climate-tracking seed dispersal is
limited in regions with relatively high climate
velocities, such as eastern North America and
Europe, as a result of past losses of large mam-
mals that provided long-distance dispersal.
Although climate-tracking dispersal is rela-
tively high in parts of Southeast Asia, this
dispersal function is threatened by current
species endangerment. Regions such as Mad-
agascar have lost substantial climate-tracking
dispersal function, and remaining function is
threatened by species endangerment. Together,
these results show that defaunation has already
limited the ability of animal-dispersed plants
in many parts of the world to keep pace with
climate change and that endangered species
provide much of the climate-tracking dis-
persal function in many other regions.
Using novel, predictive, and function-based
models of animal-mediated seed dispersal,
we show that publicly available data can be
leveraged to accurately predict species inter-
actions and map resulting ecosystem functions
globally. This shows how shifting from a species-
based to a function-based understanding of
human impacts on global biodiversity can
enable predictions for attributes of novel
communities and direct monitoring of change
in ecosystem function. Using this approach,
we found that seed dispersal function globally
has declined sharply from its natural level. Not
only have human activities caused the climate
to change rapidly—requiring broad-scale range
shifts by plants—but defaunation of birds and
mammals reduces the ability of plants to shift
their ranges. Synergistic threats to seed disperser
movement imposed by habitat fragmentation
and other land-use changes ( 34 ) will likely
amplify existing constraints on plant range
shifts. This underscores the need to not only
promote habitat connectivity to maximize
the functional potential of current seed dis-
persers ( 35 ) but also restore biotic connectivity
through the recovery of large-bodied animals
( 29 ) to increase the resilience of vegetation
communities under climate change.REFERENCESANDNOTES- P. Jordano, inSeeds: The Ecology of Regeneration in Plant
Communities(2013), pp. 18–61. - F. M. Schurret al.,Long-Distance Seed Dispersal.
L. Østergaard, Eds. (Wiley, 2018), Vol. 38, pp. 204–237. - J. P. González-Varoet al.,Nature 595 , 75–79 (2021).
- H. S. Rogers, I. Donoso, A. Traveset, E. C. Fricke,Annu. Rev.
Ecol. Evol. Syst. 52 , 641–666 (2021). - K. R. McConkeyet al.,Biol. Conserv. 146 ,1–13 (2012).
- D. H. Janzen, P. S. Martin,Science 215 , 19–27 (1982).
- C. E. Doughtyet al.,Ecography 39 , 194–203 (2016).
SCIENCEscience.org 14 JANUARY 2022•VOL 375 ISSUE 6577 213
Fig. 4. Seed dispersal tracking local climate change velocities and the impacts of defaunation and
species endangerment on climate-tracking dispersal.(A) Spatial variation in the climate-tracking
dispersal index, which describes the quantity of seeds that disperse further than the local yearly climate
velocity. (B) Loss of, and threat to, climate-tracking dispersal function caused by past defaunation and
current species endangerment are shown with a bivariate color scale. Past function lost (red axis) indicates
the percent decline in climate-tracking dispersal function when comparing current function to the natural
level. Current function endangered (blue axis) indicates the percent of currently remaining function that
would be lost if all vulnerable and endangered species were to go extinct, calculated by comparing current
and endangered species extinction scenarios.
RESEARCH | REPORTS