Content uploaded by Gonzalo Saldías
Author content
All content in this area was uploaded by Gonzalo Saldías on Jul 02, 2021
Content may be subject to copyright.
remote sensing
Article
Climatic Regulation of Vegetation Phenology in Protected Areas
along Western South America
Carlos Lara 1,2 , Gonzalo S. Saldías 3,4,* , Bernard Cazelles 5,6 , Marcelo M. Rivadeneira 7,8,9 ,
Richard Muñoz 10 , Alexander Galán 11 , Álvaro L. Paredes 12 , Pablo Fierro 13 and Bernardo R. Broitman 14
Citation: Lara, C.; Saldías, G.S.;
Cazelles, B.; Rivadeneira, M.M.;
Muñoz, R.; Galán, A.; Paredes, Á.L.;
Fierro, P.; Broitman, B.R. Climatic
Regulation of Vegetation Phenology
in Protected Areas along Western
South America. Remote Sens. 2021,13,
2590. https://doi.org/10.3390/
rs13132590
Academic Editor: Randolph H.
Wynne
Received: 6 April 2021
Accepted: 29 June 2021
Published: 2 July 2021
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
1Departamento de Ecología, Facultad de Ciencias, Universidad Católica de la Santísima Concepción,
Concepción 4090541, Chile; carlos.lara@ucsc.cl
2Centro de Investigación en Recursos Naturales y Sustentabilidad (CIRENYS), Universidad Bernardo
O’Higgins, Santiago 8370993, Chile
3Departamento de Física, Facultad de Ciencias, Universidad del Bío-Bío, Concepción 4051381, Chile
4Centro FONDAP de Investigación en Dinámica de Ecosistemas Marinos de Altas Latitudes (IDEAL),
Valdivia 5090000, Chile
5UMMISCO, UMI 209, Sorbonne Université-IRD, 75006 Paris, France; cazelles@biologie.ens.fr
6IBENS, UMR CNRS 8197, Eco-Evolution Mathématique, Ecole Normale Supérieure, 75005 Paris, France
7Centro de Estudios Avanzados en Zonas Áridas, Coquimbo 1781681, Chile; marcelo.rivadeneira@ceaza.cl
8Departamento de Biología Marina, Facultad de Ciencias del Mar, Universidad Católica del Norte,
Coquimbo 1781421, Chile
9Departamento de Biología, Universidad de La Serena, La Serena 1720256, Chile
10 Programa de Magister en Oceanografía, Universidad de Concepción, Concepción 4070386, Chile;
richmunoz@udec.cl
11 Centro de Investigación de Estudios Avanzados del Maule (CIEAM), Vicerrectoría de Investigación y
Postgrado & Departamento de Obras Civiles, Facultad de Ciencias de la Ingeniería, Universidad Católica del
Maule, Talca 3460000, Chile; agalan@ucm.cl
12 Data Observatory Foundation, Santiago 7941169, Chile; alvaro.paredes@dataobservatory.net
13 Instituto de Ciencias Marinas y Limnológicas, Facultad de Ciencias, Universidad Austral de Chile,
Valdivia 5090000, Chile; pablo.fierro@uach.cl
14 Departamento de Ciencias, Facultad de Artes Liberales and Bioengineering Innovation Center, Facultad de
Ingeniería y Ciencias, Universidad Adolfo Ibáñez, Viña del Mar 2562340, Chile; bernardo.broitman@uai.cl
*Correspondence: gsaldias@ubiobio.cl
Abstract:
Using 19 years of remotely sensed Enhanced Vegetation Index (EVI), we examined the
effects of climatic variability on terrestrial vegetation of six protected areas along southwestern
South America, from the semiarid edge of the Atacama desert to southern Patagonia (30
◦
S–51
◦
S).
The relationship between satellite phenology and climate indices, namely MEI (Multivariate ENSO
Index), PDO (Pacific Decadal Oscillation) and SAM (Southern Annular Mode) were established
using statistical analyses for non-stationary patterns. The annual mode of phenological activity
fluctuated in strength through time from the semiarid region to the border of southern Patagonia.
Concomitantly, enhanced synchrony between EVI and climatic oscillations appeared over interannual
cycles. Cross correlations revealed that variability in MEI was the lead predictor of EVI fluctuations
over scales shorter than 4 months at lower latitudes and for the most poleward study site. The PDO
was correlated with EVI over lags longer than 4 months at low latitude sites, while the SAM showed
relationships with EVI only for sites located around 40
◦
S. Our results indicate that the long-term
phenological variability of the vegetation within protected areas along southwestern South America is
controlled by processes linked to climate indices and that their influence varies latitudinally. Further
studies over longer time scales will be needed to improve our understanding the impacts of climate
change on vegetation condition and its effect over phenological variability.
Keywords: climatic change; vegetation index; MODIS; phenology; long-term variability
1. Introduction
Climatic variability has been identified as the lead driver of observed changes in
phenological cycles worldwide [
1
–
3
]. Phenological changes have consequences that cascade
Remote Sens. 2021,13, 2590. https://doi.org/10.3390/rs13132590 https://www.mdpi.com/journal/remotesensing
Remote Sens. 2021,13, 2590 2 of 17
across the entire ecosystems: from altering the timing of flower emergence, and their
mismatch with pollinators life cycles, to an earlier onset of greening and delayed senescence
of forest canopies [
4
–
7
]. Climate-driven drought disturbs the phenology of forests in boreal,
temperate, and tropical ecosystems, impacting tree growth and mortality [
8
,
9
]. On the other
hand, the strong warming-induced trend for earlier spring and later autumn has changed
the photosynthesis-mediated carbon uptake in the temperate forests of eastern North
America during the past two decades [
10
]. Although changes in phenological patterns
have been reported mostly for the northern hemisphere [
11
,
12
], the impacts of climate
variability on the dynamics of vegetation phenology have a global reach and are expected
to display a complex structure over space and time [13].
Understanding the association between climatic change and vegetation phenology
requires long–term evaluation of the spatiotemporal interplay between oceanic conditions
and land vegetation characteristics. Capturing vegetation characteristics solely based on
in situ measures is challenging due to the large spatial heterogeneity across the scales [
14
].
To circumvent this shortcoming, spectral indices derived from remotely sensed platforms
have been employed as a proxy to monitor annual and interannual variability from the
field [
15
] to regional [
16
] and global scale [
17
]. With the proliferation of satellites and
radiometric measurements, several vegetation indices have been developed among which
the Enhanced Vegetation Index (EVI) and the Normalized Difference Vegetation Index
(NDVI) have been more highlighted for their enhanced performance in capturing plant
phenology [18,19].
In hydro-climatological studies, oceanic conditions and oscillations have been pre-
sented in the form of large-scale climate indices that capture patterns of heat and mass
distribution in the atmosphere and oceans [
20
,
21
]. Among climate indices, ENSO (El Niño-
Southern Oscillation) has been recognized as one of the leading modes of global climate
variability, and other climatic modes, such as the Pacific Decadal Oscillation (PDO) or the
Atlantic Multidecadal Oscillation (AMO) interact with ENSO to modulate phenological
patterns of vegetation at a global scale [22,23].
The temporal dynamics of climate indices characterize the state of the coupled ocean—
atmosphere systems across the entire Pacific ocean, which is tightly coupled to global-scale
patterns of climatic variability, from interannual to decadal scales [24].
Using the spectral indices researchers have studied climate indices and vegetation
phenology. For example the authors of Zhao et al.
[25]
have shown ENSO (El Niño-Southern
Oscillation) as a key driver of interannual variability of vegetation phenology using NDVI.
Besides climatic effects, the influence of precipitation and temperature on phenological
activity has been studied in different regions (e.g., arid, semiarid, tropical) [
26
,
27
]. In the
semi-arid climate of the middle east, the agricultural crops are severely stressed by heat and
limited
rainfall [28]
while in Mediterranean areas, vegetation phenology is stressed during
summer due to soil moisture depletion [
29
] and in southern Africa and South America the
growth of vegetation is limited by water availability [
30
]. In temperate forest, the increase
in air temperature has also altered vegetation phenological patterns (e.g., phenological
advances of approximately 3–8 days for each
◦
C increase of 1), vegetation activity and
regional carbon cycling [10,31].
The impact of climate change is particularly important and more pronounced over
biodiversity-rich areas [
32
]. The coast of Chile, extending along western south America
from the Atacama desert to Patagonia region, spans 12 vegetation formations and 127
vegetation types [
33
]. This broad latitudinal gradient provides a model system to evaluate
the impacts of climate variability on the spatio-temporal dynamics of vegetation phenology.
Ecological studies have recognized the importance of climatic variability as a driver of
changes in the productivity and composition of the different vegetation types found along
this latitudinal gradient [
34
,
35
]. For example, long-term monitoring in semiarid central
Chile has shown that vegetation dynamics are tightly linked to interannual changes in the
ENSO-controlled precipitation regime [36].
Remote Sens. 2021,13, 2590 3 of 17
Moreover, an ENSO-driven drought now extending for over a decade (2010–2020)
in central Chile has imposed different trends on terrestrial productivity depending on
latitude [
37
]. Similarly, the coupling of positive ENSO and SAM anomalies seem to
underpin a pattern of increased forest fires across central and southern Chile over the same
period during the last decade [
21
,
38
–
40
]. Together, these studies suggest that long-term
climatic variability is impacting seems to modulate vegetation dynamics along western
south America and that their effects are not homogeneous in space neither in space nor in
time. However, the temporal climatic dynamics driving observed changes in vegetation
dynamics have not been investigated to date.
Satellite-derived vegetation indices have been used before to assess the temporal
variability of different vegetation ecosystems of Chile [
29
,
30
]. However, their ability to
capture phenological heterogeneity and their linkages with large-scale climatic oscillations
are currently lacking.
In this study, we evaluated the sensitivity of EVI to climatic variability along an exten-
sive region characterized by heterogeneous biomass distribution and hypothesized that
climatic oscillations have an asymmetric influence on the spatiotemporal phenological
variability across the study region. To this end, the aim of this study was to comprehen-
sively examine the effects of climatic variability on the EVI over multiple temporal scales,
with an emphasis on the impacts that the time-varying climate has on the phenological
patterns along the meridional extent. Section 2introduces the study area and satellite data.
Section 3describes the main results evaluating the terrestrial plant phenology response to
climatic oscillations along the latitudinal gradient. The discussion is presented in Section 4,
emphasizing the differences in phenological responses to climate change. Finally, the main
conclusions are highlighted in Section 5.
2. Data and Methodological Analysis
2.1. Study Area
The central–southern coast of Chile along the western Pacific coast of South America
is characterized by a narrow semiarid region, where the Atacama desert, the driest on the
planet, changes into a Mediterranean climate (30
◦
S–37
◦
S). While humidity in the semiarid
region is provided mostly by seasonal coastal fog, the Mediterranean region is characterized
by seasonal precipitation (
>
200 mm) concentrated during the austral winter months (June,
July, August), which is subject to considerable interannual variability. On the other hand,
the southern–austral region experiences a temperate climate (38
◦
S–53
◦
S), with year-round
rainfall that can exceed 2000 mm, particularly south of 40◦S [41–43].
The vegetation of the lower-latitude part of the coastal temperate sector, between
38 and 43◦S
, corresponds to Valdivian rainforest, dominated by species with wide leaves
adapted to high humidity [
44
]. The Sub-Antarctic forest poleward of 47
◦
S is dominated
by deciduous Nothofagus species such as Nothofagus pumilio and Nothofagus antarctica,
among
others [44,45]
. Due to the high endemism and its isolation, the Valdivian and Sub-
Antarctic forests integrate a global biodiversity hotspot [
46
,
47
], which has been affected by
human disturbances (selective logging and livestock grazing) and natural processes [48].
We focused our study on forested areas located within the boundaries of National
Parks to tease out the potential effects of long-term human activities on vegetation (i.e.,
changes in land use and/or irrigation) from climate-driven phenological dynamics. In Chile,
the conservation strategy that defines the legal boundaries and management plans for these
protected areas follows that stated by the International Union for Conservation of Nature
(IUCN) as: “a clearly defined geographical space, recognized, dedicated and managed,
through legal or other effective means, to achieve the long-term conservation of nature
with associated ecosystem services and cultural values”. These areas generally include
heterogeneous vegetation and land surface cover, for example rocky or glaciated areas,
and are managed under the National System of Protected Wild Areas (SNASPE), which is
administrated by the Chilean National Forestry Corporation (CONAF) [
33
,
49
]. We selected
six national parks that spanned the vegetation gradient from the semi-arid edge of the
Remote Sens. 2021,13, 2590 4 of 17
Mediterranean sector to the Cold Patagonian tundra, using cartography provided by the
SNASPE to define the units of study (Figure 1). The selected parks were: Fray Jorge (30
◦
S),
La Campana (32
◦
S), Alerce Costero (40
◦
S), Tantauco (43
◦
S), Laguna San Rafael (47
◦
S),
and Torres del Paine (51◦S) (Table 1).
Table 1.
Vegetation formations in National Parks of this study according to Pliscoff and Fuentes-
Castillo [33].
National Park Area (km2)Year Created Vegetation Formation
(1) Fray Jorge (30◦S) 86.06 1941 Desertic scrub
(2) La Campana (32◦S) 78.18 1967 Deciduous forest
Sclerophyllous forest
Alpine dwarf scrub
(3) Alerce Costero (40◦S) 236.14 2010 Valdivian Laured-
Leaved Forest
Fitzroya Forest
(4) Tantauco (43◦S) 1016.4 2005 Deciduous forest
(5) San Rafael (47◦S) 18,693.40 1959 Deciduous forest
Sclerophyllous forest
Evergreen forest
Alpine dwarf scrub
Deciduous shrub
Evergreen shrub
Peatbog
(6) Torres del Paine (51◦S) 2216.47 1959 Deciduous forest
Evergreen forest
Alpine dwarf scrub
Matorral arborescent
Deciduous shrub
Remote Sens. 2021,13, 2590 5 of 17
Figure 1.
Map of the study area showing National Parks: (
a
) Fray Jorge, (
b
) La Campana, (
c
) Alerce
Costero, (d) Tantauco, (e) San Rafael and, (f) Torres del Paine.
2.2. Phenology and Climatic Data
We used satellite images of the Enhanced Vegetation Index (EVI; MOD13Q1 product,
version 6) from the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor
onboard the Terra satellite. EVI is an L3 product with a resolution of 250 m, generated
every 16 days from daily images. The EVI algorithm choose the best available pixel
value (i.e., low clouds, low view angle, and the highest NDVI/EVI value) from all the
acquisitions from the 16 days period. We downloaded and analyzed a time series of images
from February 2000 to December 2018. Using the polygon of each site, we generated an
internal 250-m buffer for each park and the corresponding EVI time series (e.g., [
50
]). Our
analysis aimed to capture biospheric responses to climate, and hence they included the
totality of the park area and the different kinds of vegetation within (see Table 1). Zonal
statistics per observation (scene) were calculated for each park, at each time. This results in
a temporal series of one data point every 16 days, for each park (the mean). That data were
aggregated by month using a weighted mean, in order to match the temporal resolution of
the climatic data, with the weight consisting in the number of valid pixels per scene at each
time and park. When there was only one scene per month, that observation is used. When
there was two, the weighted mean is computed. All data extracting and processing were
implemented in Python and the MODIS imagery re-projection and cropping were carried
out using the Modis Reprojection Tool.
In order to identify the long-term relationships between climatic oscillations and
phenological variability, we examined three climatic indices: MEI, PDO and SAM, which
were obtained from the National Weather Service website (www.cpc.ncep.noaa.gov, ac-
cessed on 1 September 2020) of the National Oceanic and Atmospheric Administration
(NOAA). These three monthly climatic indices strongly influence climate along western
south America, and their variability modulates changes in precipitation and streamflow
along the study region [51,52].
Remote Sens. 2021,13, 2590 6 of 17
2.3. Data Analysis
We performed a wavelet analysis on MODIS phenological data to explore the interan-
nual variability in phenological indices. Additionally, a bootstrapping method was used to
evaluate the significance of wavelet spectra (more details are found in [
53
]). The relative
importance of frequencies for each time step is represented in the time–frequency plane
to form the local wavelet power spectrum (WPS). Here, we used the Morlet wavelet to
detect the characteristics in phenology fluctuations, which according to Cazelles et al.
[54]
is defined as:
ψ(t) = π−1/4ex p(−iω0t)exp(−t2/2)(1)
where
ex p(−iω0t)
is the product of a complex sinusoidal,
π−1/4
is a normalization factor
to ensure unit variance and
ω0
represent the central angular frequency of the
wavelet [54]
.
One of the key advantages of a continuous wavelet function is the mathematical relationship
between the wavelet scale and its frequency
a
and its frequency
f
. We also computed
the global WPS where all wavelet coefficients of the same frequency
f
were averaged.
The global WPS is defined by Cazelles et al. [54] as:
WPSx(f) = σ2
x
TZT
0kWx(f,τ)k2dτ(2)
where
σ2
x
is the variance of time-series (EVI),
T
is the time duration of the time series (2000–
2018) and
Wx(f
,
τ)
is the wavelet transform of the
x(t)
. All analyses were performed using
the MATLAB wavelet scripts (www.biologie.ens.fr/~cazelles/bernard/Research.html,
accessed on 6 April 2021). To quantify the co-variation between EVI and climate indices we
computed the wavelet coherence, which measures the correlation between the spectra of
two series [55] and is defined as:
Rx,y(f,τ) = k<Wx,y(f,τ)>k
k<Wx,x(f,τ)>k1/2k<Wy,y(f,τ)>k1/2 (3)
where
<>
indicates a smoothing operator in time and scale,
Wx,y(f
,
τ)
is the wavelet
co–spectrum of the two time series
x(t)
and
y(t)
, respectively,
Wx,x(f
,
τ)
and
Wy,y(f
,
τ)
are
the wavelet transform of the
x(t)
and
y(t)
. The wavelet coherence (
Rx,y(f
,
τ)
) is equal to
1 when there is a perfect linear relation between the frequency of two temporal signals,
and is equal to 0 when signals
x(t)
and
y(t)
are independent with no common frequencies
at a particular temporal scale [
54
,
56
]. Finally, we computed cross-correlations between
the EVI signal and the climatic indices for each park with time lags from 0 to 7 months.
The semi-annual lag was chosen as the maximum delay where the impact of climate-driven
seasonal variability could be detected on the phenological cycle.
3. Results
Vegetation productivity, using EVI as proxy, showed a significant interannual variabil-
ity in the intensity of the 1-year period—the phenological band. In general, WPS of EVI
revealed a unique and significant annual signal, that persisted throughout the study period
for all evaluated parks (Figure 2), except for La Campana and Torres del Paine, where this
phenological signal was interrupted. Changes in EVI were significant just during 2001–2006,
2010–2013 and 2015–2017 for La Campana (Figure 2b) and during 2001–2009 and 2012–2018
for Torres del Paine, being the last period characterized by low WPS
(Figure 2f)
. Another
low-powered but significant oscillation of about 2 year was also detected during 2006–2009
and 2016–2017 at this sub-antarctic region (Figure 2f). This shifting in the phenological
signal intensity across the latitudinal range suggests an external time-varying forcing of
the annual cycle of productivity.
Remote Sens. 2021,13, 2590 7 of 17
Figure 2.
Wavelet analysis of EVI monthly data between 2000 and 2018 in (
a
) Fray Jorge, (
b
) La
Campana, (
c
) Alerce Costero, (
d
) Tantauco, (
e
) San Rafael and (
f
) Torres del Paine. The left panels
show local wavelet power spectrum. The color code for power values is from white (low power
values) to dark red (high power values). The dot-black lines indicate the 95% significant areas
obtained by adapted bootstrapping [
57
] and the cone of influence (bold-black lines) indicates the
wavelet domain where computations are not influenced by edge effects (see [
56
]). The right panels
show the global wavelet power spectrum for the period examined (2000–2018) with the black line
showing a 95% confidence interval obtained by adaptive bootstrapping (note that they have different
scale for better visualization).
Figure 3shows the WPS for all climate indices studied. While the MEI showed a
high-power significant oscillation just for the 2–3 year period between 2007 and 2016
(
Figure 3a
), the PDO denoted a transient and significant 1.5–2 year oscillation during 2001–
2003 and 2007–2013, with another significant oscillation of 2–4 years for the 2013–2016
period (
Figure 3b
). Conversely, high-power spectra oscillation for SAM were less frequent
relative to MEI and PDO, with significant WPS occurring around a 6–7 year band for the
2010–2016 period (Figure 3c).
Transient and significant synchrony between EVI and MEI was revealed over multi-
year frequencies (1 to 4-year) for the different parks. A significant and persistent 2–3-year
component was present between 2004 and 2011 (Fray Jorge, Tantauco and Torres del
Paine), and 2003 and 2006 (La Campana and Alerce Costero), while a significant 4-year
periodic component was observed between 2008 and 2016 (San Rafael), 2010 and 2015 (La
Campana), and
2012 and 2015
(Alerce Costero) (Figure 4). The synchrony between EVI
and PDO showed a more consistent pattern across time and space. From 2010 to 2016,
a periodic 1-year oscillation was observed along the entire studied area, while a transient
and broad synchronous band over a 2–3-year period was observed between 2001 and 2005
and 2008 and 2015 in Fray Jorge,
2001 and 2013
in Tantauco and 2002 and 2011 in Torres del
Paine (Figure 5). The temporal coherence between EVI and SAM revealed a 2–4 year joint
band in Fray Jorge (2008–2013), La Campana (2006–2015), Alerce Costero (2002–2009) and
Torres del Paine (2002–2007), and a 4–6 year oscillation appeared significant in Tantauco
(2009–2015) and San Rafael (2003–2007) (Figure 6). In several portions of the analysis,
intra-anual (0.5-year) modes were significant, yet they should be interpreted with caution
due to their short duration and the coarse temporal resolution of our data.
Remote Sens. 2021,13, 2590 8 of 17
Figure 3.
Wavelet analysis of climatic indices monthly data between 2000 and 2018. (
a
) MEI,
(b) PDO
and (
c
) SAM. The left panels show local wavelet power spectrum. The color code for power values is
from white (low power values) to dark red (high power values). The dot-black lines indicate the 95%
significant areas obtained by adapted bootstrapping [
57
] and the cone of influence (bold-black lines)
indicates the wavelet domain where computations are not influenced by edge effects (see [
56
]). The
right panels show the global wavelet power spectrum for the period examined (2000–2018) with the
black line showing a 95% confidence interval obtained by adaptive bootstrapping (note that they
have different scale for better visualization).
Figure 4.
Wavelet coherence between the EVI and MEI temporal signal in national parks. (
a
) Fray
Jorge, (
b
) La Campana, (
c
) Alerce Costero, (
d
) Tantauco, (
e
) San Rafael and (
f
) Torres del Paine.
The colors are coded from yellow (low coherence) to dark red (high coherence). The dotted-dashed
blue lines show the 95% and the 90% significance levels computed based on bootstrapped series.
The cone of influence (black line) indicates the region not influenced by edge effects (see [56]).
Remote Sens. 2021,13, 2590 9 of 17
Figure 5.
Wavelet coherence between the EVI and PDO temporal signal in national parks. (
a
) Fray Jorge,
(b) La Campana
,
(c) Alerce Costero
, (
d
) Tantauco, (
e
) San Rafael and (
f
) Torres del Paine. The colors are coded from yellow (low coherence)
to dark red (high coherence). The dotted-dashed blue lines show the 95% and the 90% significance levels computed based
on bootstrapped series. The cone of influence (black line) indicates the region not influenced by edge effects (see [56]).
Figure 6.
Wavelet coherence between the EVI and SAM temporal signal in national parks. (
a
) Fray Jorge,
(b) La Campana
,
(c) Alerce Costero
, (
d
) Tantauco, (
e
) San Rafael and (
f
) Torres del Paine. The colors are coded from yellow (low coherence)
to dark red (high coherence). The dotted-dashed blue lines show the 95% and the 90% significance levels computed based
on bootstrapped series. The cone of influence (black line) indicates the region not influenced by edge effects (see [56]).
Lagged correlations between climatic indices and EVI are presented in Figure 7.
Climate indices lead the analysis and positive correlations mean that the corresponding
climate index and EVI are in phase (e.g., positive MEI before leads to positive EVI after).
At Fray Jorge Park, EVI showed a quick and weakly significant positive relationship
with MEI that persisted for 3 months. For PDO, the response was negative at lag-0 and
switched to a positive relationship over lags between 6 and 7 months. No significant
cross-correlations were observed with the SAM at this site (Figure 7a). A similar pattern
was observed in La Campana National Park, but the positive cross-correlation between
EVI and MEI persisted for 5 months and the positive significant lags with PDO appeared
at 6 and
7 months
(Figure 7b). Alerce Costero and Tantauco National Parks, located ca.
10 degrees south of La Campana, did not share the lagged correlation structure with the
Mediterranean region. No significant correlations with MEI were observed, but there
were significant negative associations with PDO between 2- and 5-month lags for Alerce
Costero and 4-month lag for Tantauco, which also showed a significant and positive zero-
Remote Sens. 2021,13, 2590 10 of 17
lag correlation with SAM (Figure 7c,d, respectively). In the case of San Rafael National
Park, significant and positive correlations were observed at lag-0 for both PDO and SAM,
and also for SAM at
1 month
(Figure 7e), while for Torres del Paine, the most austral
National Park, EVI showed significant positive correlations with MEI at 3–4 months and
negative at 0 and 4–5 months with PDO (Figure 7f).
Figure 7.
Lagged correlations between EVI and MEI (blue line), EVI and SAM (orange line) and EVI
and PDO (green line) in each park (see order in Figure 1). The external numbers around the ring
show the respective lag (0 to 7) between the climatic data and EVI. Points marked with asterisk and
on the light gray zone of the circle are statistically significant at 5%.
4. Discussion
Our work shows a persistent and time-varying association between phenological
patterns and different climatic oscillations over a broad range of vegetational formations
spread along the western coast of south America. These results provide new insights on
how the spatio-temporal variable natural climate forcing modulates the physiological status
in terrestrial ecosystems over global, regional and local scales (see references
in [58,59])
and influences the provision of essential environmental variables.
Due to the critical role that vegetation plays in the global biogeochemical and hydro-
logical cycles, and the services we derive from them, it is relevant to understand how plant
dynamics respond to global climatic oscillations over multiples scales [
60
]. Our study high-
lights the impacts that tropical ocean–atmosphere dynamics have on seasonal vegetation
Remote Sens. 2021,13, 2590 11 of 17
activity, ranging from the semi-arid Mediterranean scrub down to the sub Antarctic tundra
of Patagonia. Previous studies evidence the strong impact of ENSO events (e.g., extreme
MEI anomalies) on vegetation of the Southern Hemisphere [
34
,
51
]. The warm phase of
ENSO (El Niño) covaries with severe drought events in some regions (e.g., Australia, south-
east Asia, northeastern South America), whereas the cold phase (La Niña) is associated
with droughts in western south America [
22
,
61
]. In concomitance with ENSO, PDO also
appears as a driver of the phenological patterns along western south America, mediated by
the meridional displacement of the subtropical gyre, changes in the pattern of the winds,
and SST anomalies [43,62,63].
Changes in the precipitation pattern associated with negative phases of PDO are
related to a drying trend in central Chile (30
◦
S–40
◦
S) over the last 30 years [
42
,
43
,
64
].
The detrimental effects of the reduced rainfall are evidenced as changes in the vege-
tation growing-season during the so-called megadrought period (2010–2018) in central
Chile [37,64]
. The uninterrupted sequence of dry years, i.e., years with a rainfall deficit,
coincided with a negative EVI signal over pasture lands and areas with native forests cover.
When coupled with climate effects (e.g., warming), these intense drought periods could
promote fire activity (frequency and severity), mainly in areas with extensive plantations
of non-native forests, native sclerophyll forests, and scrublands, vegetation types that
largely determine the primary productivity patterns in central Chile [
37
,
65
]. While in
northern Patagonia, both coastal ocean and land biological processes are forced by climatic
variability over multiple temporal scales [66].
In the Southern Hemisphere, where instrumental records are sparse and relatively
short, the opportunities for understanding the large-scale dynamical mechanisms driving
climate variability are limited [
67
]. According to Marshall
[68]
, the SAM oscillation is
defined by the gradient between the high-pressure belt at medium latitudes and the lower
pressures around coastal Antarctica. As such, it is calculated as the zonally averaged
pressure difference between 40
◦
S and 65
◦
S. SAM oscillations are phase-dependent with
ENSO, e.g., when a positive phase of ENSO (El Niño) coincide with a negative phase
of SAM (Figure A2), climate anomalies dominate south of 40
◦
S [
69
], generating extreme
conditions due to high solar irradiation and reduced precipitation [
66
,
70
]. Such phase-
locking could drive part of the observed changes in phenological activity, e.g., weak
amplitude (see Figure 7).
Despite the low but significant lagged-effect of ENSO and PDO on the phenological
pattern for the 43
◦
S–52
◦
S zone, our results strengthen the conclusion that the teleconnec-
tion between ocean—atmosphere oscillations are evident in the coast of South America.
Multiple studies have associated ENSO with spring rainfall in southern China [
71
] and
across western North America [
72
]. Although these climatic drivers have not been formally
established, following a lack of in situ data, our results can be used to assess the impacts of
climatic teleconnection processes with local phenology patterns.
EVI provides a consistent measure of greenness and improves sensitivity over dense
vegetation conditions closely related to the seasonal variation of biological activity and
their sensitivity to climatic oscillations over multiple time-lag relationships [
73
]. Through-
out the study period (2000–2018), we observed significant time-lagged effects of the local
vegetation to global climatic oscillations. However, in Fray Jorge National Park, several
long-term studies that investigated the effects of climatic forces on population dynamics of
wild rodents suggest that pulses of higher rainfall, induced by ENSO, impact ecological
processes over multiple scales of organizations (e.g., [
74
–
76
]), including human communi-
ties and land use. Garreaud et al.
[37]
showed the effects of climate variability on rainfall
along a latitudinal gradient spanning most of Chile and including all types of vegetation.
It showed that, along the central region (30–38
◦
S), a large negative rainfall anomaly (nick-
named Megadrought) had a major impact on the gross plant productivity (as measured
by the EVI index) during the 2010–2015 period, disproportionally impacted the semi-arid
shrublands of north–central Chile (30–33
◦
S) and the coastal region south of 30
◦
S, following
the curtailed rainfall during this period. In contrast, positive but patchy anomalies in
Remote Sens. 2021,13, 2590 12 of 17
EVI were recorded between 34
◦
S and 37
◦
S, which were associated to forest plantations
involving non-native species with a rapid initial growth in leaf area [
77
]. Thus, the com-
plex interplay among climate-driven variability in air temperature, rainfall, and other
variables in the latitudinal gradient of Chile, in addition to the different vegetation types
(
Table 1
), should result in marked differences in phenological activity (evidenced in the
power spectra for different parks; Figures 2and A1) in our study. Therefore, it is essential to
couple the impact of climate variability with the patterns of air temperature, precipitations,
and hydrological regimes to further inspect the contrasting impacts of climate events on
the plant productivity and phenology along western South America.
Remote sensing vegetation indices can be used as a robust tool to assist decision-
making for the conservation and management of areas with high vegetation endemism.
Time-lagged responses provide us with insights into how earlier conditions impact vege-
tation activity. In most open shrubs, at middle and low latitudes, climatic forcing reveals
their significant effect on vegetation growth [
73
]. This study highlights the complexity in
biogeographic climate–vegetation biomass effects that should be considered in predictive
climate models of phenological variations along western South-America.
5. Summary
The Pacific coast of western South America has a wide latitudinal range of climate
variability, with consequent change in vegetation communities. In recent years there has
been increasing interest regarding the effects of climatic oscillations on phenological pat-
terns. Our study demonstrated that the response of EVI to capture the annual cycle of
vegetative communities in the temperate climate region minimizes canopy—soil variations
and improves sensitivity over dense vegetation conditions. On the other hand, the MEI
and SAM indices showed an effect on multiple lags along the latitudinal gradient. Co-
incidentally, PDO also presents an effect (4–6 month lags) along the study area. Future
studies should carefully analyze the PDO as the ENSO/SAM variability only partially
determines the phenological patterns along our broad study region. To this end, our results
provide a first step towards the design of conservation strategies that integrate climatic
information into the managing and monitoring of essential environmental variables, chiefly
the phenology terrestrial productivity, along the Pacific coast of western South America.
Author Contributions:
C.L., B.C., M.M.R., Á.L.P. and B.R.B. conceived the research; C.L., B.C.,
M.M.R., G.S.S., A.L.P. working in data curation and formal analysis; C.L., B.C., M.M.R., G.S.S., R.M.,
A.G., Á.L.P., P.F. and B.R.B. writing the original draft and editing the final version. All authors have
read and agreed to the published version of the manuscript.
Funding:
The LP DAAC operates as a partnership between the U.S. Geological Survey (USGS) and
the National Aeronautics and Space Administration (NASA), for the production and distribution
of Terra Moderate Resolution Imaging Spectroradiometer (MODIS) Vegetation Indices (MOD13Q1).
This study was supported by FONDECYT grant 11190209 to CL, 1200843 to MMR, 1190805 to GSS,
11190631 and 1181300 to BRB. This work was partially supported by the Millennium Institute for
Coastal Socio-ecology (SECOS) and the Millenium Nucleus Understanding Past coastal upWelling
system and Environmental Local and Lasting impacts (UPWELL), Millenium Science Initiative
Programs—ICN2019015 and NCN19153, respectively, and the CYTED program under the grant
number 520RT0010 Red GeoLIBERO—Consolidación de una red de geomática libre aplicada a las
necesidades de Iberoamérica. Maria José Martínez-Harms provided insightful comments to an earlier
draft, for which we are most grateful.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement:
The vegetation indices from the MODIS-TERRA data product were
retrieved from the online Data Pool, courtesy of the NASA Land Processes Distributed Active
Archive Center (LP DAAC), USGS/EarthResources Observation and Science (EROS) Center, Sioux
Falls, South Dakota (https://lpdaac.usgs.gov/products/mod13q1v006/, accessed on 6 April 2021).
Climatic indices are available in National Weather Service website (www.cpc.ncep.noaa.gov, accessed
on 6 April 2021) of the National Oceanic and Atmospheric Administration (NOAA).
Remote Sens. 2021,13, 2590 13 of 17
Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design
of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript,
or in the decision to publish the results.
Appendix A
In this contribution, we also showed the NDVI (Normalized Difference Vegetation
Index)’s temporal trends (Figure A1). As EVI, the NDVI also serves for detecting temporal
changes in vegetation phenological activity. EVI have a greater dynamic range than NDVI
to capture the phenological cycle, due to the fact that NDVI is mainly saturated in the
red band region of the electromagnetic region where the energy is strongly absorbed by
pigments [
78
,
79
] and also these values strongly tend to decrease in the presence of high
background amounts of dry biomass. Moreover, NDVI varies depending the chlorophyll—
the concentration and amount of vegetation biomass saturating when both chlorophyll
and biomass present high levels [
78
,
80
]. For example, in the Chiloé Island Ecosystem,
recognized as a biodiversity “hotspot”, NDVI is susceptible to uncertainty of capturing the
annual photosynthetic
cycle [79]
. EVI perform well under high aerosol loads (which may
add significant noise to the signal), biomass burning conditions and topographic effect [
78
].
However, the temporal co-evolution of EVI and NDVI showed in general complementary
patterns across the study region. In terms of the effects of climatic variability over both EVI
and NDVI, our results showed the asymmetrical response of the phenological signal.
Figure A1.
Wavelet analysis of NDVI monthly data between 2000 and 2018 in (
a
) Fray Jorge,
(b) La Campana
, (
c
) Alerce Costero, (
d
) Tantauco, (
e
) San Rafael and (
f
) Torres del Paine. The
left panels show the local wavelet power spectrum. The color code for power values is from white
(low power values) to dark red (high power values). The dot-black lines indicate the 95% significant
areas obtained by adapted bootstrapping [
57
] and the cone of influence (bold-black lines) indicates
the wavelet domain where computations are not influenced by edge effects (see [
56
]). The right
panels show the global wavelet power spectrum for the period examined (2000–2018), with the black
line showing a 95% confidence interval obtained by adaptive bootstrapping.
Remote Sens. 2021,13, 2590 14 of 17
Figure A2.
Time series of Climate indices: Multivariate ENSO Index (MEI; red/blue), Southern
Annular Mode (SAM; black line), and Pacific Decadal Oscillation (PDO; grey line).
References
1.
White, M.A.; Hoffman, F.; Hargrove, W.W.; Nemani, R.R. A global framework for monitoring phenological responses to climate
change. Geophys. Res. Lett. 2005,32. [CrossRef]
2.
Donnelly, A.; Caffarra, A.; O’Neill, B.F. A review of climate-driven mismatches between interdependent phenophases in terrestrial
and aquatic ecosystems. Int. J. Biometeorol. 2011,55, 805–817. [CrossRef] [PubMed]
3.
Renner, S.S.; Zohner, C.M. Climate change and phenological mismatch in trophic interactions among plants, insects, and
vertebrates. Annu. Rev. Ecol. Syst. 2018,49, 165–182. [CrossRef]
4.
Cleland, E.E.; Chuine, I.; Menzel, A.; Mooney, H.A.; Schwartz, M.D. Shifting plant phenology in response to global change.
Trends Ecol. Evol. 2007,22, 357–365. [CrossRef] [PubMed]
5.
Inouye, D.W. Effects of climate change on phenology, frost damage, and floral abundance of montane wildflowers. Ecology
2008
,
89, 353–362. [CrossRef]
6.
Seidl, R.; Thom, D.; Kautz, M.; Martin-Benito, D.; Peltoniemi, M.; Vacchiano, G.; Wild, J.; Ascoli, D.; Petr, M.; Honkaniemi, J.; et al.
Forest disturbances under climate change. Nat. Clim. Chang. 2017,7, 395–402. [CrossRef]
7.
Beard, K.H.; Kelsey, K.C.; Leffler, A.J.; Welker, J.M. The missing angle: Ecosystem consequences of phenological mismatch. Trends
Ecol. Evol. 2019,34, 885–888. [CrossRef] [PubMed]
8.
Allen, C.D.; Macalady, A.K.; Chenchouni, H.; Bachelet, D.; McDowell, N.; Vennetier, M.; Kitzberger, T.; Rigling, A.;
Breshears, D.D.;
Hogg, E.T.; et al. A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests.
For. Ecol. Manag. 2010,259, 660–684. [CrossRef]
9.
Klos, R.J.; Wang, G.G.; Bauerle, W.L.; Rieck, J.R. Drought impact on forest growth and mortality in the southeast USA: An analysis
using Forest Health and Monitoring data. Ecol. Appl. 2009,19, 699–708. [CrossRef]
10.
Keenan, T.F.; Gray, J.; Friedl, M.A.; Toomey, M.; Bohrer, G.; Hollinger, D.Y.; Munger, J.W.; O’Keefe, J.; Schmid, H.P.; Wing, I.S.; et al.
Net carbon uptake has increased through warming-induced changes in temperate forest phenology. Nat. Clim. Chang.
2014
,
4, 598–604. [CrossRef]
11.
Piao, S.; Fang, J.; Zhou, L.; Ciais, P.; Zhu, B. Variations in satellite-derived phenology in China’s temperate vegetation. Glob.
Chang. Biol. 2006,12, 672–685. [CrossRef]
12.
Ma, Q.; Huang, J.G.; Hänninen, H.; Berninger, F. Reduced geographical variability in spring phenology of temperate trees with
recent warming. Agric. For. Meteorol. 2018,256–257, 526–533. [CrossRef]
13.
Yuan, H.; Wu, C.; Gu, C.; Wang, X. Evidence for satellite observed changes in the relative influence of climate indicators on
autumn phenology over the Northern Hemisphere. Glob. Planet Chang. 2020,187, 103131. [CrossRef]
14.
Maignan, F.; Bréon, F.M.; Bacour, C.; Demarty, J.; Poirson, A. Interannual vegetation phenology estimates from global AVHRR
measurements: Comparison with in situ data and applications. Remote Sens. Environ. 2008,112, 496–505. [CrossRef]
15.
Jamshidi, S.; Zand-Parsa, S.; Niyogi, D. Assessing Crop Water Stress Index of Citrus Using In-Situ Measurements, Landsat, and
Sentinel-2 Data. Int. J. Remote Sens. 2021,42, 1893–1916. [CrossRef]
16.
Wang, X.; Zhang, S.; Feng, L.; Zhang, J.; Deng, F. Mapping Maize Cultivated Area Combining MODIS EVI Time Series and the
Spatial Variations of Phenology over Huanghuaihai Plain. Appl. Sci. 2020,10, 2667. [CrossRef]
17.
Ye, W.; van Dijk, A.I.; Huete, A.; Yebra, M. Global trends in vegetation seasonality in the GIMMS NDVI3g and their robustness.
Int. J. Appl. Earth Obs. Geoinf. 2021,94, 102238. [CrossRef]
18.
Kowalski, K.; Senf, C.; Hostert, P.; Pflugmacher, D. Characterizing spring phenology of temperate broadleaf forests using Landsat
and Sentinel-2 time series. Int. J. Appl. Earth Obs. Geoinf. 2020,92, 102172. [CrossRef]
Remote Sens. 2021,13, 2590 15 of 17
19.
Sarvia, F.; De Petris, S.; Borgogno-Mondino, E. Exploring Climate Change Effects on Vegetation Phenology by MOD13Q1 Data:
The Piemonte Region Case Study in the Period 2001–2019. Agronomy 2021,11, 555. [CrossRef]
20.
Gore, M.; Abiodun, B.J.; Kucharski, F. Understanding the influence of ENSO patterns on drought over southern Africa using
SPEEDY. Clim. Dyn. 2020,54, 307–327. [CrossRef]
21.
Sarricolea, P.; Serrano-Notivoli, R.; Fuentealba, M.; Hernández-Mora, M.; de la Barrera, F.; Smith, P.; Meseguer-Ruiz, Ó. Recent
wildfires in Central Chile: Detecting links between burned areas and population exposure in the wildland urban interface. Sci.
Total Environ. 2020,706, 135894. [CrossRef]
22.
Woodward, F.; Lomas, M.; Quaife, T. Global responses of terrestrial productivity to contemporary climatic oscillations. Philos.
Trans. R. Soc. B Biol. Sci. 2008,363, 2779–2785. [CrossRef]
23.
Ito, A. Decadal variability in the terrestrial carbon budget caused by the Pacific Decadal Oscillation and Atlantic Multidecadal
Oscillation. J. Meteorol. Soc. Ser. II 2011,89, 441–454. [CrossRef]
24.
Dätwyler, C.; Grosjean, M.; Steiger, N.J.; Neukom, R. Teleconnections and relationship between the El Niño–Southern Oscillation
(ENSO) and the Southern Annular Mode (SAM) in reconstructions and models over the past millennium. Clim. Past
2020
,
16, 743–756. [CrossRef]
25.
Zhao, L.; Dai, A.; Dong, B. Changes in global vegetation activity and its driving factors during 1982–2013. Agric. For. Meteorol.
2018,249, 198–209. [CrossRef]
26.
Ficetola, G.F.; Maiorano, L. Contrasting effects of temperature and precipitation change on amphibian phenology, abundance and
performance. Oecologia 2016,181, 683–693. [CrossRef] [PubMed]
27.
Liu, Q.; Fu, Y.H.; Zeng, Z.; Huang, M.; Li, X.; Piao, S. Temperature, precipitation, and insolation effects on autumn vegetation
phenology in temperate China. Glob. Chang. Biol. 2016,22, 644–655. [CrossRef] [PubMed]
28.
Jamshidi, S.; Zand-Parsa, S.; Niyogi, D. Physiological responses of orange trees subject to regulated deficit irrigation and partial
root drying. Irrig. Sci. 2021,39, 441–455. [CrossRef]
29.
Gordo, O.; Sanz, J.J. Impact of climate change on plant phenology in Mediterranean ecosystems. Glob. Chang. Biol.
2010
,
16, 1082–1106. [CrossRef]
30.
Zhang, X.; Hodges, J.C.; Schaaf, C.B.; Friedl, M.A.; Strahler, A.H.; Gao, F. Global vegetation phenology from AVHRR and MODIS
data. In Proceedings of the IGARSS 2001, Scanning the Present and Resolving the Future, IEEE 2001 International Geoscience
and Remote Sensing Symposium (Cat. No. 01CH37217), Sydney, Australia, 9–13 July 2001; Volume 5, pp. 2262–2264.
31.
Richardson, A.D.; Keenan, T.F.; Migliavacca, M.; Ryu, Y.; Sonnentag, O.; Toomey, M. Climate change, phenology, and phenological
control of vegetation feedbacks to the climate system. Agric. For. Meteorol. 2013,169, 156–173. [CrossRef]
32.
Bhattacharjee, A.; Anadón, J.D.; Lohman, D.J.; Doleck, T.; Lakhankar, T.; Shrestha, B.B.; Thapa, P.; Devkota, D.; Tiwari, S.; Jha, A.;
et al. The impact of climate change on biodiversity in Nepal: Current knowledge, lacunae, and opportunities. Climate
2017
,5, 80.
[CrossRef]
33.
Pliscoff, P.; Fuentes-Castillo, T. Representativeness of terrestrial ecosystems in Chile’s protected area system. Environ. Conserv.
2011,38, 303–311. [CrossRef]
34.
Jaksic, F.M. Ecological effects of El Niño in terrestrial ecosystems of western South America. Ecography
2001
,24, 241–250.
[CrossRef]
35.
Squeo, F.A.; Loayza, A.P.; López, R.P.; Gutiérrez, J.R. Vegetation of Bosque Fray Jorge National Park and its surrounding matrix
in the Coastal Desert of north-central Chile. J. Arid Environ. 2016,126, 12–22. [CrossRef]
36.
Holmgren, M.; Scheffer, M.; Ezcurra, E.; Gutiérrez, J.R.; Mohren, G.M. El Niño effects on the dynamics of terrestrial ecosystems.
Trends Ecol. Evol. 2001,16, 89–94. [CrossRef]
37.
Garreaud, R.D.; Alvarez-Garreton, C.; Barichivich, J.; Boisier, J.P.; Christie, D.; Galleguillos, M.; LeQuesne, C.; McPhee, J.;
Zambrano-Bigiarini, M. The 2010–2015 megadrought in central Chile: Impacts on regional hydroclimate and vegetation. Hydrol.
Earth Syst. Sci. 2017,21, 6307–6327. [CrossRef]
38. Úbeda, X.; Sarricolea, P. Wildfires in Chile: A review. Glob. Planet Change 2016,146, 152–161. [CrossRef]
39.
Urrutia-Jalabert, R.; González, M.E.; González-Reyes, Á.; Lara, A.; Garreaud, R. Climate variability and forest fires in central and
south-central Chile. Ecosphere 2018,9, e02171. [CrossRef]
40.
Holz, A.; Paritsis, J.; Mundo, I.A.; Veblen, T.T.; Kitzberger, T.; Williamson, G.J.; Aráoz, E.; Bustos-Schindler, C.; González, M.E.;
Grau, H.R.; et al. Southern Annular Mode drives multicentury wildfire activity in southern South America. Proc. Natl. Acad. Sci.
USA 2017,114, 9552–9557. [CrossRef]
41.
Viale, M.; Garreaud, R. Orographic effects of the subtropical and extratropical Andes on upwind precipitating clouds. J. Geophys.
Res. Atmos. 2015,120, 4962–4974. [CrossRef]
42.
Boisier, J.P.; Rondanelli, R.; Garreaud, R.D.; Muñoz, F. Anthropogenic and natural contributions to the Southeast Pacific
precipitation decline and recent megadrought in central Chile. Geophys. Res. Lett. 2016,43, 413–421. [CrossRef]
43.
Quintana, J.; Aceituno, P. Changes in the rainfall regime along the extratropical west coast of South America (Chile): 30–43
º
S.
Atmósfera 2012,25, 1–22.
44. Alaniz, A.J. Chilean forest: Recent trends and perspectives. Persp. Chall. 2019, 1–308.
45.
Madariaga, G.C.; Navarro, R.C.; Baeza, Á.P.; Verrugio, G.C. Structural and biometric characterization of Nothofagus betuloides
production forests in the Magellan Region, Chile. J. Sustain. For. 2007,24, 123–140. [CrossRef]
Remote Sens. 2021,13, 2590 16 of 17
46.
Myers, N.; Mittermeier, R.A.; Mittermeier, C.G.; Da Fonseca, G.A.; Kent, J. Biodiversity hotspots for conservation priorities.
Nature 2000,403, 853. [CrossRef]
47.
Díaz, I.A.; Armesto, J.J.; Reid, S.; Sieving, K.E.; Willson, M.F. Linking forest structure and composition: Avian diversity in
successional forests of Chiloé Island, Chile. Biol. Conserv. 2005,123, 91–101. [CrossRef]
48.
Echeverría, C.; Newton, A.C.; Lara, A.; Benayas, J.M.R.; Coomes, D.A. Impacts of forest fragmentation on species composition
and forest structure in the temperate landscape of Southern Chile. Glob. Ecol. Biogeogr. 2007,16, 426–439. [CrossRef]
49.
Petit, I.J.; Campoy, A.N.; Hevia, M.J.; Gaymer, C.F.; Squeo, F.A. Protected areas in Chile: Are we managing them? Rev. Chil. Hist.
Nat. 2018,91, 1. [CrossRef]
50.
Jusys, T. Quantifying avoided deforestation in Pará: Protected areas, buffer zones and edge effects. J. Nat. Conserv.
2016
,33, 10–17.
[CrossRef]
51.
Núñez, J.; Rivera, D.; Oyarzún, R.; Arumí, J. Influence of Pacific Ocean multidecadal variability on the distributional properties
of hydrological variables in north-central Chile. J. Hydrol. 2013,501, 227–240. [CrossRef]
52.
Saldías, G.S.; Largier, J.L.; Mendes, R.; Pérez-Santos, I.; Vargas, C.A.; Sobarzo, M. Satellite-measured interannual variability of
turbid river plumes off central-southern Chile: Spatial patterns and the influence of climate variability. Prog. Oceanogr.
2016
,
146, 212–222. [CrossRef]
53.
Cazelles, B. Symbolic dynamics for identifying similarity between rhythms of ecological time series. Ecol. Lett.
2004
,7, 755–763.
[CrossRef]
54.
Cazelles, B.; Chavez, M.; Magny, G.C.d.; Guégan, J.F.; Hales, S. Time-dependent spectral analysis of epidemiological time-series
with wavelets. J. R. Soc. Interface 2007,4, 625–636. [CrossRef] [PubMed]
55.
Chatfield, C.; Xing, H. The Analysis of Time Series: An Introduction with R; CRC Press, Taylor & Francis Group: Boca Raton, FL,
USA, 2019.
56.
Cazelles, B.; Chavez, M.; Berteaux, D.; Ménard, F.; Vik, J.O.; Jenouvrier, S.; Stenseth, N.C. Wavelet analysis of ecological time
series. Oecologia 2008,156, 287–304. [CrossRef]
57.
Cazelles, B.; Cazelles, K.; Chavez, M. Wavelet analysis in ecology and epidemiology: Impact of statistical tests. J. R. Soc. Interface
2014,11, 20130585. [CrossRef]
58.
Morisette, J.T.; Richardson, A.D.; Knapp, A.K.; Fisher, J.I.; Graham, E.A.; Abatzoglou, J.; Wilson, B.E.; Breshears, D.D.;
Henebry, G.M.;
Hanes, J.M.; et al. Tracking the rhythm of the seasons in the face of global change: Phenological research
in the 21st century. Front. Ecol. Environ. 2009,7, 253–260. [CrossRef]
59.
Taig-Johnston, M.; Strom, M.K.; Calhoun, K.; Nowak, K.; Ebensperger, L.A.; Hayes, L. The ecological value of long-term studies
of birds and mammals in Central America, South America and Antarctica. Rev. Chil. Hist. Nat. 2017,90, 1–13. [CrossRef]
60.
Buermann, W.; Anderson, B.; Tucker, C.J.; Dickinson, R.E.; Lucht, W.; Potter, C.S.; Myneni, R.B. Interannual covariability in
Northern Hemisphere air temperatures and greenness associated with El Niño-Southern Oscillation and the Arctic Oscillation.
J. Geophys. Res. Atmos. 2003,108. [CrossRef]
61.
Trenberth, K.E.; Dai, A.; Van Der Schrier, G.; Jones, P.D.; Barichivich, J.; Briffa, K.R.; Sheffield, J. Global warming and changes in
drought. Nat. Clim. Chang. 2014,4, 17–22. [CrossRef]
62.
Chen, W.; Feng, J.; Wu, R. Roles of ENSO and PDO in the link of the East Asian winter monsoon to the following summer
monsoon. J. Clim. 2013,26, 622–635. [CrossRef]
63.
Schneider, W.; Donoso, D.; Garcés-Vargas, J.; Escribano, R. Water-column cooling and sea surface salinity increase in the upwelling
region off central-south Chile driven by a poleward displacement of the South Pacific High. Prog. Oceanogr.
2017
,151, 38–48.
[CrossRef]
64.
Garreaud, R.D.; Boisier, J.P.; Rondanelli, R.; Montecinos, A.; Sepúlveda, H.H.; Veloso-Aguila, D. The central Chile mega drought
(2010–2018): A climate dynamics perspective. Int. J. Climatol. 2020,40, 421–439. [CrossRef]
65.
McWethy, D.B.; Pauchard, A.; García, R.A.; Holz, A.; González, M.E.; Veblen, T.T.; Stahl, J.; Currey, B. Landscape drivers of recent
fire activity (2001–2017) in south-central Chile. PLoS ONE 2018,13, e0201195. [CrossRef]
66.
Lara, C.; Cazelles, B.; Saldías, G.S.; Flores, R.P.; Paredes, Á.L.; Broitman, B.R. Coupled Biospheric Synchrony of the Coastal
Temperate Ecosystem in Northern Patagonia: A Remote Sensing Analysis. Remote Sens. 2019,11, 2092. [CrossRef]
67.
Hessl, A.; Allen, K.J.; Vance, T.; Abram, N.J.; Saunders, K.M. Reconstructions of the southern annular mode (SAM) during the last
millennium. Prog. Phys. Geogr. 2017,41, 834–849. [CrossRef]
68.
Marshall, G.J. Trends in the Southern Annular Mode from observations and reanalyses. J. Clim.
2003
,16, 4134–4143. [CrossRef]
69. Gong, D.; Wang, S. Definition of Antarctic oscillation index. Geophys. Res. Lett. 1999,26, 459–462. [CrossRef]
70. Garreaud, R. Record-breaking climate anomalies lead to severe drought and environmental disruption in western Patagonia in
2016. Clim. Res. 2018,74, 217–229. [CrossRef]
71.
Wu, X.; Mao, J. Interdecadal modulation of ENSO-related spring rainfall over South China by the Pacific Decadal Oscillation.
Clim. Dyn. 2016,47, 3203–3220. [CrossRef]
72.
Schoennagel, T.; Veblen, T.T.; Romme, W.H.; Sibold, J.S.; Cook, E.R. ENSO and PDO variability affect drought-induced fire
occurrence in Rocky Mountain subalpine forests. Ecol. Appl. 2005,15, 2000–2014. [CrossRef]
73.
Wu, D.; Zhao, X.; Liang, S.; Zhou, T.; Huang, K.; Tang, B.; Zhao, W. Time-lag effects of global vegetation responses to climate
change. Glob. Chang. Biol. 2015,21, 3520–3531. [CrossRef]
Remote Sens. 2021,13, 2590 17 of 17
74.
Gutierrez, J.R.; Meserve, P.L.; Kelt, D.A.; Engilis Jr, A.; Previtali, M.A.; Milstead, W.B.; Jaksic, F.M. Long-term research in Bosque
Fray Jorge National Park: Twenty years studying the role of biotic and abiotic factors in a Chilean semiarid scrubland. Rev. Chil.
Hist. Nat. 2010,83, 69–98. [CrossRef]
75.
Meserve, P.L.; Kelt, D.A.; Previtali, M.A.; Milstead, W.B.; Gutiérrez, J.R. Global climate change and small mammal populations in
north-central Chile. J. Mammal. 2011,92, 1223–1235. [CrossRef]
76.
Meserve, P.L.; Kelt, D.A.; Gutiérrez, J.R.; Previtali, M.A.; Milstead, W.B. Biotic interactions and community dynamics in the
semiarid thorn scrub of Bosque Fray Jorge National Park, north-central Chile: A paradigm revisited. J. Arid Environ.
2016
,
126, 81–88. [CrossRef]
77.
Le Maire, G.; Marsden, C.; Nouvellon, Y.; Grinand, C.; Hakamada, R.; Stape, J.L.; Laclau, J.P. MODIS NDVI time-series allow the
monitoring of Eucalyptus plantation biomass. Remote Sens. Environ. 2011,115, 2613–2625. [CrossRef]
78.
Huete, A.; Didan, K.; Miura, T.; Rodriguez, E.P.; Gao, X.; Ferreira, L.G. Overview of the radiometric and biophysical performance
of the MODIS vegetation indices. Remote Sens. Environ. 2002,83, 195–213. [CrossRef]
79.
Lara, C.; Saldías, G.S.; Paredes, A.L.; Cazelles, B.; Broitman, B.R. Temporal variability of MODIS phenological indices in the
temperate rainforest of Northern Patagonia. Remote Sens. 2018,10, 956. [CrossRef]
80.
Mutanga, O.; Skidmore, A.K. Narrow band vegetation indices overcome the saturation problem in biomass estimation. Int. J.
Remote Sens. 2004,25, 3999–4014. [CrossRef]