Untitled
Untitled
Untitled
https://doi.org/10.1007/s10457-023-00821-7 (0123456789().,-volV)
( 01234567
89().,-volV)
Abstract Homegarden (HG) agroforestry combines the arboreal species. Homegardens, regardless of
biological carbon (C) sequestration with biodiversity elevation or size, contribute to C sequestration and
conservation outcomes. Although C stocks and agrobiodiversity conservation and help achieve the
species richness of HGs vary along elevational UN Sustainable Development Goals (SDGs), partic-
gradients and as a function of holding sizes, there is ularly Climate Action (SDG-13) and conserving
no consensus on the nature and magnitude of such agrobiodiversity (SDG-15, Life on Land).
variations. Field studies were conducted in the
Western Ghats region of central Kerala, India (180 Keywords Agrobiodiversity Carbon
homesteads in 20 selected panchayats), to evaluate the sequestration Climate change mitigation Western
effects of elevation (near sea level to 1938 m) and Ghats Sustainable development goals
garden size (162–10,117 m2) on aboveground C stocks
and floristic diversity. The C stocks (per unit area) of
HGs (arborescent species) were highly variable
(0.63–93.65 Mg ha–1), as garden management was Introduction
highly individualistic and it exhibited a weak negative
relationship with elevation. Likewise, there was a Agroforestry systems (AFS) contribute significantly to
weak negative relationship between C stocks and the global carbon (C) cycle due to their high biomass
garden size. Tree stocking levels (stems/garden) and production and C accumulation potentials (Nair et al.
species richness (species/garden) positively impacted 2021). Homegarden agroforestry (Fig. 1) not only has
total C stocks per garden. Floristic diversity was high a large potential for biological C sequestration in soil
in the study area (753 species) and included many rare and biomass (Kumar and Kunhamu 2021) but also is
and endangered species (43 IUCN Red-Listed species) an icon of circa situm biodiversity conservation
making homegardens circa situm reservoirs of biodi- (preserving planted and/or relic trees and wildings
versity. Elevation and holding size exerted a weak on farmlands where forests or woodlands of similar
negative linear relationship on Simpson’s floristic composition existed once; Dawson et al. 2013; Kumar
diversity index, which ranged from 0.26 to 0.93 for 2011). Distributed predominantly in the tropical areas
of Asia, Africa, Central America, the Caribbean, and
the Pacific Islands (Nair et al. 2021), the homegardens
B. M. Kumar (&)
Arunachal University of Studies, Knowledge City,
(HGs) are rich in plant diversity and have been ranked
Namsai, Arunachal Pradesh 792103, India first among all man-made agroecosystems in terms of
e-mail: bmohankumarkau@gmail.com
123
Agroforest Syst
123
Agroforest Syst
reported that for southern Ethiopian HGs, total Materials and methods
biomass and soil C stocks were significantly
(p = 0.001) higher in the upper (1877–2000 m) Field surveys were conducted in central Kerala, India,
elevation zone than in the mid (1754–1876 m) and at four locations, namely, Thrissur, Ernakulum,
lower (1630–1753 m) zones with 26% more C stock Idukki, and Malappuram districts (Fig. 2). The study
in the upper zone. Bargali and Bargali (2020) also area falls within the broad biogeographic region of the
showed that soil C stocks increased with increasing Western Ghats, a ‘‘global biodiversity hotspot’’, and
altitude (350–2000 m) for small HGs in the Central experiences a warm humid tropical climate. The
Himalayas, India; but for medium and large HGs, it principal soil types, number of households, and other
increased up to the mid-altitude and decreased details of the study locations are described in KSLUB
thereafter, signifying a complex interrelationship (1995). HGs constitute the predominant land use
between C stocks, size of gardens, and altitudinal activity in the study area (Kumar and Nair 2004;
positions. Kumar and Takeuchi 2009; Kumar 2011), which,
The association between biological C sequestra-
tion and plant diversity has become a subject of
huge scientific interest in the discourse on climate
change mitigation. Although this hypothesis on
species diversity and net primary production (i.e.,
species-diverse systems are more productive) was
tested for natural systems (Tilman et al. 1997), very
little research has been conducted to validate these
claims in an agroforestry setting. Saha et al. (2009)
reported that the soil C stock was directly related to
the plant diversity of HGs. Yet, the interrelation-
ships between species richness and aboveground
biomass C stocks were seldom evaluated for HGs. It
was, therefore, hypothesized that an increase in the
number of species may influence the accumulation
of biomass C and as a result, HGs that have higher
species diversity may have greater C sequestration
potential (CSP). However, operational attributes
such as the size of HGs and managerial interven-
tions may have a complex effect on species richness
and biomass productivity. Therefore, it was further
hypothesized that although biomass C stocks of
Kerala HGs may co-vary with species richness along
an elevational gradient, operational attributes such
as the size of the HGs and managerial interventions
may have an overriding impact on species diversity
and CSP. The specific objectives of the study were
(1) to investigate how the aboveground C stocks and
species diversity attributes of Kerala HGs varied
along an elevational gradient, (2) to examine the
differences in aboveground C stocks in HGs of
varying sizes, and (3) to assess the interrelationships
between the aboveground C stocks and species
Fig. 2 Map of Kerala showing the study locations. Shapefile
diversity/richness. reprinted from GADM database under CC BY license
123
Agroforest Syst
however, are being cleared for various purposes, but was measured and the number of culms per clump was
offset by the expansion of younger, less diverse counted, besides measuring clump height using a Haga
agroforests into paddy wetlands (Fox et al. 2017). Altimeter. Within each HG, three randomly selected
For estimating carbon stocks and species diversity 1 m 9 1 m quadrats were also laid out, and the
of the HGs along an altitudinal gradient (near sea level species-wise counts (number of plants within the
to 1938 m elevation), 20 panchayats (the lowest level quadrat) of the herbaceous plants (\ 5 cm GBH) were
of local self-government) from the coastal plains to the recorded. Sample specimens of plant species that
High Ranges of Kerala were selected considering their could not be identified in the field were collected for
elevations and ease of access (i.e., 11% of all comparison with the herbaria collections maintained
panchayats in the designated districts). The survey at Kerala Agricultural University.
was conducted during the period from February 2019
to March 2020 (until the onset of the Covid-19 Computations and statistical analysis of the data
pandemic). The contact details of the farmers and
information on the approximate size of holdings and The aboveground tree biomass was computed using the
the general land-use pattern were obtained from the following allometric equations for individual trees and
Krishibhavans (agricultural development offices) of summed up to derive the garden-wise biomass esti-
the respective panchayats. The respondents were mates. However, species-specific allometric equations
contacted and a detailed investigation of their home- have not been developed for several dicot species.
steads was conducted, which included gathering Therefore, a generalized allometric equation for the
details on the nature and kind of species, socioeco- moist tropical forests (FAO 2004) was used in such
nomic parameters, and farm management. cases.
Three broad size categories of HGs viz., small
• Ailanthus triphysa: lnB ¼ 2:06622 þ
(\ 0.4 ha holding size), medium (0.4–0.8 ha holding
2:120555lnDBH; where B = Aboveground bio-
size), and large ([ 0.8 ha holding size) were included
mass, DBH = Diameter at breast height (cm)
in the study (at least three gardens each per category
(Kumar et al. 1998).
and panchayath) and a total of 180 HGs with a total
• Artocarpus heterophyllus: InB ¼ 1:71891þ
farm area of 190.18 ha was sampled. The extent of
2:251298lnDBH, where B = Aboveground bio-
HGs within the total farm was ascertained from the
mass, DBH = Diameter at breast height (cm)
gardeners, and it was occasionally cross-checked by
(Kumar et al. 1998).
field observations. In many small gardens, however,
• Artocarpus hirsutus: lnB ¼ 3:06934þ
the boundaries of the HGs and the total farms were co-
2:793477lnDBH, where B = Aboveground bio-
terminus with one another. Geographic coordinates
mass, DBH = Diameter at breast height (cm)
and altitudes of each HG surveyed were noted using a
(Kumar et al. 1998).
Garmin eTrex 30 9 GPS (USA). Mean monthly
• Emblica officinalis (syn. Phyllanthus emblica):
temperature and precipitation data for the study
LnB ¼ 2:70511 þ 0:969805lnðDBH 2 HÞ, where
locations for the 1991–2021 period were downloaded
B = Aboveground biomass, DBH = Diameter at
from the Climate-data.org portal (Table 1; https://en.
breast height (cm), H = Total height (m) (Kumar
climate-data.org/asia/india/kerala-752/).
et al. 1998).
All arboreal species (trees and shrubs: ] 5 cm
girth at breast height; 1.37 m, GBH) present in the • Hevea brasiliensis: B ¼ 0:002604G2:7826 , where
HGs, including the border rows, were enumerated by B = Shoot weight in kg and G = girth in cm at
measuring their GBH, bole height, and total height 150 cm height (Shorrocks et al. 1965).
using a measuring tape and Haga Altimeter, respec- • Tectona grandis:
tively. For plantation species such as coconut (Cocos Log10 B ¼ 0:815 þ 2:382log10 DBH, where
nucifera), rubber (Hevea brasiliensis), areca nut B = Aboveground biomass, DBH = Diameter at
(Areca catechu), and nutmeg (Myristica fragrans), breast height (cm) (Kraenzel et al. 2003).
10 random plants were measured, in addition to • All other dicot trees: Y¼
recording their total numbers as stated by the expf2:134 þ 2:53lnðDÞg; where Y = biomass
landowner. For bamboos, GBH of 10 random culms per tree in kg, exp = exponential function,
123
Agroforest Syst
Table 1 Latitude, longitude, and elevation of the study areas (Panchayats) and the range, mean, and standard error values of carbon
stocks of homegarden trees and shrubs (] 5 cm GBH) in central Kerala, India with meteorological parameters of the study locations
Panchayat Latitude Longitude Elevation Homegarden biomass carbon Mean Mean annual
(N) (E) (m) stocks (Mg ha–1) monthly precipitation* (mm)
temperature*
(C)
Min Max Min Max Mean Standard Min Max
Error
Adimali 10.02 76.93 529 867 24.49 64.58 36.7 4.72 20.8 24.1 2672
Alagappanagar 10.44 76.29 19 36 24.88 73.54 46.4 5.98 24.9 28.3 3161
Athirapilly 10.30 76.49 34 102 21.82 74.52 40.5 6.36 24.2 28.1 3161
Chakkupallam 9.77 77.05 1031 1146 27.40 68.53 41.4 4.81 21.3 24.6 1548
Chelakkara 10.69 76.34 33 67 27.15 68.21 46.7 4.37 24.6 28.9 2242
Erumapetty 10.68 76.19 17 38 36.03 71.97 58.1 3.97 24.9 28.8 2746
Kandanassery 10.51 76.08 14 37 12.46 75.16 47.7 4.88 25.3 28.6 2522
Kavalangad 10.04 76.76 35 131 16.99 56.08 35.1 3.43 23.7 27.0 3410
Kodassery 10.34 76.39 14 45 26.04 64.79 42.1 4.79 25.4 28.5 2641
Madakkathara 10.42 76.21 23 36 25.12 66.75 36.1 4.10 25.2 28.9 2520
Orumanayur 10.55 76.04 5 11 14.00 49.86 34.5 3.54 25.3 28.6 2522
Pallivasal 10.02 76.91 708 933 7.98 83.03 31.6 7.47 19.4 22.7 2672
Pariyaram 10.31 76.42 16 177 22.88 62.37 34.6 3.98 24.5 27.6 2979
Pudukad 10.40 76.30 11 23 18.11 60.93 36.7 4.56 25.0 28.7 2520
Thiruvilwamala 10.73 76.45 50 79 19.39 66.39 39.5 5.84 24.3 29.0 2284
Vadakkekad 10.65 76.01 8 14 12.61 93.65 40.7 8.53 25.4 28.9 2696
Vallikkunnu 11.11 75.85 13 52 23.94 63.25 34.6 4.54 25.5 28.9 2746
Vandiperiyar 9.59 77.08 741 1067 22.05 68.17 46.5 5.31 20.8 23.9 3343
Vattavada 10.18 77.25 1632 1938 0.63 25.81 6.3 4.88 17.8 20.1 2932
Velur 13.35 76.16 9 35 16.42 89.51 37.3 7.73 25.1 29.0 2248
Overall mean – – 38.7 –
*Source: https://en.climate-data.org/asia/india/kerala-752/; temperature and precipitation values correspond to the 1991–2021 period.
For the Panchayats for which meteorological data were not available, the data from the nearby location available were used; for
latitude and longitude, the mean values for all homegardens surveyed are presented
D = diameter at breast height in cm. This equation • Bambusa spp. (culms): Log B = 1:2471þ
predicts the trunk and canopy biomass of moist 3:5238logD, where B is the total dry weight (kg)
(1500–4000 mm rainfall) forest trees (D \ 80 cm) and D is the culm diameter at breast height (cm)
with reasonable precision (R2 = 0.97) (FAO (Shanmughavel and Francis 1996).
2004).
Carbon sequestration was estimated based on
• Coconut palms: Y ¼ 5:5209x þ 89:355, where
biomass stocks, i.e., 50% of the total aboveground
Y = dry weight (kg); x = palm age (years) (Kumar
biomass stock was taken as C stock. Total C stock (per
and Russell, unpublished data, 2006).
garden) was divided with the HG area, which includes
• All other palms (areca palm or Areca catechu L.,
the area used for herbaceous crop production in
palmyra palm or Borassus flabellifer L., etc.):
association with trees and the portions used for
Y ¼ 4:5 þ 7:7H, where Y = biomass (kg) and
residential purposes, to derive the C stock per hectare.
H = stem height (m) (Brown 1997).
123
Agroforest Syst
Simpson’s diversity index (D) was calculated as also highlight the high degree of variability in the C
S
Pni 2 stocks within the panchayats. Elevational gradient
D¼1 N , where s = total number of spe- generally had a negative linear relationship with C
i¼1
cies, ni = number of individuals of species ‘i’, stocks (Fig. 3A). But the R2 value was low, indicating
N = total number of individuals in the plot, and D high ‘‘noise’’ levels in the dataset (F = 1.924;
is the diversity index (Simpson 1949) and Shan- p = 0.167). The size of the HGs (Fig. 3B) also did not
non index (H’) was computed as H 0 ¼ 3:3219 exert any robust influence on biomass C stocks.
Although a higher C stocks in smaller HGs was
P S
log10 N N1 ni log10 ni ; where H’ = Shannon postulated, the data presented in Fig. 3B, show only a
i¼1
index, s = total number of species, ni = number of weak negative trend (F = 2.32; p = 0.129). The size of
gardens, however, was a major determinant of the
individuals of species ‘i’, and N = total number of
number of stems of arboreal species per garden, which
individuals in the plot (Shannon and Weaver 1949). In
addition, Hmax, the maximum dispersion taking into increased more or less linearly (F = 18.78;
p \ 0.00002) and the larger HGs generally had more
account the number of species present
stems per garden (Fig. 4A). The number of species also
(Hmax ¼ 3:3219 log10 S, where S is the total number
increased with the size of the garden (Fig. 4B) in a log-
of species) and equitability (E), a measure of homo-
linear manner. The coefficients of determination for the
geneity of relative diversity (E ¼ H 0 =Hmax ), were
prediction equations were, however, low. The wide
computed. Species richness or accumulation (number
variability among the HGs in species richness (3–38
of species) and stocking levels (number of stems per
species) and the number of stems per garden (21–1640)
garden) of arboreal species (] 5 cm GBH) were
presumably may explain this lack of tight relationships
derived from the HG survey data.
Linear regression equations linking aboveground C
stocks (per hectare) with elevational positions and size A 100
of the holdings, tree stocking levels (stems per garden) y = 41.7453 – 0.00879x
Carbon stocks (Mg C/ha)
y = 42.4422 – 0.00108x
80 R² = 0.0129
Results 60
40
Aboveground C stocks
20
The aboveground C stock ranged from 0.63 to
0
93.65 Mg ha–1 for the HGs studied (Fig. 3A), implying
0 2000 4000 6000 8000 10000
great variability in the C sequestration potential of HGs. Holding size (m2)
Average standing stocks of C panchayat-wise (for all
holding sizes) were in the range of 6–58 Mg ha–1 Fig. 3 Aboveground carbon stocks (per ha) of trees and shrubs
(Table 1). The moderately high standard error values (]5 cm girth at breast height) of homegardens along an
elevational gradient (A) and as a function of garden size (B)
123
Agroforest Syst
Number of stems per garden
1800 90
(Mg)
1000 y = 121.2626 + 0.0389x 50
800 R² = 0.0954 40
600 30
400 20
200 10
0 0
0 2000 4000 6000 8000 10000 12000 0 500 1000 1500 2000
Holding size (m2) Number of stems per garden
90
Species richness (per garden)
40
(Mg)
20 40
15 30
10 20
5 10
0 0
0 2000 4000 6000 8000 10000 12000 0 10 20 30 40
Holding size (m2) Number of species per garden
Fig. 4 A: The number of stems per garden (]5 cm GBH) and Fig. 5 Aboveground biomass carbon stock as a function of (A)
B: Number of species per garden (]5 cm GBH) in the the number of stems per garden (]5 cm GBH) and (B) the
homegardens of central Kerala as a function of the size of number of species per garden (] 5 cm GBH) in the homegar-
holdings dens of central Kerala, India
among the variables. The number of stems per garden, in number of stems or species per garden (data not
turn, determined the total C stocks per garden (Fig. 5A) presented).
and the linear function linking the variables explained
the trend reasonably well with an R2 value of 47% Floristic richness and diversity
(F = 154.92; p \ 0.0001) signifying that large gardens
with a greater number of stems will have higher total C A total of 753 plant species (313 herbs, 155 shrubs,
stocks. The total C stock also increased with the number and 285 trees; Table 2 and Appendix 1) were
of species (Fig. 5B; R2 = 15%; F = 32.42; encountered in the HGs studied, which involved 43
p \ 0.0001). Aboveground C stocks per hectare, how- endangered species (8 Near Threatened, 9 Endan-
ever, did not show a distinctive trend in respect of the gered, 3 Critically Endangered, and 23 Vulnerable
species) as per the IUCN Red List (IUCN 2022).
Table 2 Summary of plant species encountered in the homegardens of central Kerala with the IUCN status
Habit Number of Species IUCN categories*
DD LC NT EN CR VU NE
123
Agroforest Syst
Table 3 List of Critically Endangered, Endangered, Near Threatened and Vulnerable plants observed in the homegardens of central
Kerala
Sl. no Species name Local name Habit IUCN—Red List category
123
Agroforest Syst
Table 3 continued
Sl. no Species name Local name Habit IUCN—Red List category
Table 4 Floristic diversity indexes (mean ± standard error) of arborescent plants ([ 5 cm GBH) in the homegardens of central
Kerala, India
Panchayat No. of stems No. of Simpson’s index Shannon index
per garden species per
garden CD D H’ Hmax E
Adimali 244 ± 33.4 15 ± 0.91 0.451 ± 0.059 0.549 ± 0.059 1.343 ± 0.144 2.668 ± 0.069 0.504 ± 0.052
Alagappanagar 134 ± 25.9 22 ± 1.44 0.166 ± 0.031 0.834 ± 0.031 2.342 ± 0.183 3.087 ± 0.071 0.752 ± 0.046
Athirapilly 118 ± 24.6 20 ± 1.80 0.204 ± 0.030 0.797 ± 0.030 2.166 ± 0.102 2.984 ± 0.083 0.728 ± 0.036
Chakkupallam 126 ± 15.5 21 ± 2.74 0.163 ± 0.014 0.837 ± 0.014 2.332 ± 0.117 2.996 ± 0.126 0.778 ± 0.020
Chelakkara 170 ± 43.8 20 ± 1.73 0.249 ± 0.042 0.751 ± 0.042 1.959 ± 0.146 2.947 ± 0.102 0.672 ± 0.053
Erumapetty 198 ± 47.9 21 ± 1.41 0.234 ± 0.045 0.766 ± 0.045 2.056 ± 0.196 3.028 ± 0.077 0.676 ± 0.057
Kandanassery 279 ± 76.3 21 ± 1.73 0.307 ± 0.047 0.693 ± 0.047 1.719 ± 0.135 3.003 ± 0.084 0.577 ± 0.047
Kavalangad 212 ± 49.3 20 ± 1.83 0.317 ± 0.053 0.683 ± 0.053 1.730 ± 0.140 2.947 ± 0.088 0.600 ± 0.063
Kodassery 376 ± 152.6 19 ± 1.77 0.364 ± 0.048 0.636 ± 0.048 1.557 ± 0.152 2.913 ± 0.099 0.530 ± 0.041
Madakkathara 319 ± 44.9 25 ± 2.69 0.230 ± 0.024 0.770 ± 0.024 2.023 ± 0.135 3.157 ± 0.117 0.636 ± 0.027
Orumanayur 169 ± 39.7 17 ± 1.17 0.327 ± 0.049 0.673 ± 0.049 1.607 ± 0.171 2.816 ± 0.065 0.563 ± 0.049
Pallivasal 644 ± 169.0 24 ± 1.58 0.314 ± 0.054 0.687 ± 0.054 1.701 ± 0.145 3.150 ± 0.069 0.547 ± 0.054
Pariyaram 102 ± 13.5 19 ± 2.17 0.161 ± 0.014 0.839 ± 0.014 2.222 ± 0.115 2.888 ± 0.122 0.767 ± 0.021
Pudukad 193 ± 33.7 19 ± 1.84 0.227 ± 0.019 0.773 ± 0.019 1.945 ± 0.095 2.902 ± 0.096 0.671 ± 0.027
Thiruvilwamala 83 ± 8.1 19 ± 1.17 0.200 ± 0.031 0.800 ± 0.031 2.197 ± 0.111 2.923 ± 0.063 0.749 ± 0.027
Vadakkekad 221 ± 63.7 21 ± 2.52 0.249 ± 0.026 0.751 ± 0.026 1.900 ± 0.124 3.008 ± 0.121 0.637 ± 0.045
Vallikkunnu 223 ± 53.1 23 ± 2.92 0.301 ± 0.067 0.699 ± 0.067 1.877 ± 0.199 3.053 ± 0.123 0.618 ± 0.064
Vandiperiyar 308 ± 58.5 26 ± 1.48 0.249 ± 0.028 0.751 ± 0.028 2.011 ± 0.098 3.240 ± 0.059 0.621 ± 0.029
Vattavada 38 ± 5.8 5 ± 0.75 0.343 ± 0.048 0.657 ± 0.048 1.258 ± 0.155 1.472 ± 0.165 0.855 ± 0.033
Velur 266 ± 77.5 18 ± 2.09 0.357 ± 0.055 0.643 ± 0.055 1.584 ± 0.193 2.833 ± 0.133 0.562 ± 0.067
GBH Girth at breast height (1.37 m), CD Concentration of dominance, D Simpson’s diversity index, H’ Shannon diversity index,
Hmax = Maximum diversity, E Equitability or evenness
Significantly, this included as many as 20 vulnerable locational attributes in determining stocking levels
trees, which are listed in Table 3 along with the other (number of stems per garden) and species richness of
endangered species. the HGs. For example, a high-altitude location such as
The number of stems and species per HG showed Vattavada (near Munnar), where winter veg-
considerable variability among the studied gardens etable growing is the predominant occupation of the
and their locations (Table 4), implying the interplay of villagers, had the smallest suite of species and
holding size and management factors, besides stocking levels (five species and 38 stems per
123
Agroforest Syst
1.00 Discussion
Simpson's Divesity Index (D)
0.80
Homegardens for climate action
0.60
Biological C sequestration, a key strategy for reducing
0.40
y = 0.74497 – 0.000053x atmospheric concentrations of CO2 and for mitigating
R² = 0.0048
0.20 global warming, is a central feature of HGs. A
A comparison of the data in Table 1 indicates that the
0.00
0 2000 4000 6000 8000 10000 12000 aboveground biomass C stocks of HGs in central
Holding size (m2) Kerala (arboreal species) ranged between 6 and
58 Mg ha–1. The high variability among the panchay-
Simpson's Diversity Index (D)
1.00
0.90 ats may be because of the interplay of a complex set of
0.80
factors such as species composition of the gardens
0.70
0.60 (Table 4), HG sizes which ranged from 162 to 10,117
0.50 m2, site characteristics, tree age, tree stocking or
0.40 density, and tree/land management practices (Nair
0.30
0.20 y = 0.735895 – 0.000021x et al. 2010; Kumar 2011). Many of these parameters
0.10 R² = 0.0041 differed widely among the gardens and among the
B
0.00 physiographic zones of Kerala (midlands, highlands,
0 500 1000 1500 2000
and coastal plains). Regardless of such variations, the
Elevaonal gradient (m)
aboveground standing C stocks were generally very
Fig. 6 A Simpson’s Diversity Index of plants (] 5 cm girth at high ([ 30 Mg ha–1). Kumar (2011) in the previously
breast height) as a function of the size of homegardens and mentioned study also reported average aboveground
B along an elevational gradient standing stocks of C in the range of 16–36 Mg ha-1
for trees in the size class above 20 cm GBH, which is
broadly consistent with the present observations. The
garden respectively). Conversely, Panchayats like high biomass C stocks indicate the potential of HGs in
Madakkathara, Pallivasal, and Vandiperiyar located climate change mitigation strategies (e.g., SDG-13;
in vastly different altitudinal zones (Table 1) showed a Sharma et al. 2022). For instance, in neighboring Sri
greater number of species (24 to 26 per garden) and Lanka, HGs figure prominently in the Intended
stems (] 5 cm GBH) per garden (308 to 644 stems), Nationally Determined Contributions (INDCs) to the
signifying the highly location-specific and unique United Nations Framework Convention on Climate
nature of the HG systems. Change (UNFCCC) and in meeting the climate
The Simpson’s diversity index (D) of the sampled mitigation goals under UN REDD ? Program to cut
HG trees and shrubs (] 5 cm GBH; Table 4) ranged emissions from deforestation and forest degradation
from 0.55 (Adimali) to 0.84 (Pariyaram) and the (Mattsson et al. 2018). Nonetheless, locations like
density-based Shannon index (H’) from 1.34 (Adi- Vattavada (Munnar), where winter vegetable growing
mali) to 2.34 (Alagappanagar), with Hmax values of is the predominant land use activity in the managed
1.472 (Vattavada) to 3.24 (Vandiperiyar), signifying ecosystems, had low C stock values such as
moderately high species diversity. The moderate to 6 Mg ha-1, which highlights the crucial role that land
high equitability or evenness index, 0.504 (Adimali) to use management plays in determining biomass C
0.855 (Vattavada), indicates that the stems were more stocks.
or less evenly distributed among the species. Neither
holding size (Fig. 6A; F = NS) nor elevation (Fig. 6B; Homegarden C stocks along the elevational
F = NS) revealed any distinct floristic diversity pat- gradient
tern for the arboreal species, but a faint negative
correlation was visible for both parameters. Although elevation is generally considered an impor-
tant determinant (negative) of biomass C stocks,
elevation of HGs showed only a weak negative
123
Agroforest Syst
correlation in the present study (Fig. 3A). For natural managerial interventions, and the structural attributes
ecosystems, many previous studies, however, have of HGs play a cardinal role in determining HG
reported robust negative relationships between eleva- biomass C stocks than elevation per se or size of the
tion and C stocks (Raich et al. 2006; Leuschner et al. holdings.
2007; Malhi et al. 2017). The underlying functional Large gardens with a greater number of stems per
relationship between elevational positions and forest garden (Fig. 4A) had higher total C stocks (per
net primary productivity is usually explained by the garden), as expected (Fig. 5A). Species richness also
strong correlation between elevation and mean annual exerted a favorable influence on total C stocks
temperature (Girardin et al. 2013) and/or precipitation (Fig. 5B). High species assemblages in the HGs may
seasonality (Cruz-Amo et al. 2020). The variations in promote higher net primary productivity as they
mean monthly temperature values (range: 17.8–29.0 harbor divergent species presumably with contrasting
C; Table 1) and precipitation seasonality (a general resource-utilization characteristics, in turn, promoting
bimodal distribution pattern with the mean annual C sequestration, compared to less species-rich sys-
precipitation ranging from 1548 mm for Chakkupal- tems. Thompson et al. (2012) reported that the C stock
lam to 3410 mm for Vandiperiyar, about 20 km way; in a system depends on both functional diversity and
Table 1) for the present study locations, however, were the diversity of woody species. Therefore, a positive
not profound, implying that other resource gradients relationship between plant species richness and C
that co-vary with elevation, and more importantly, the sequestration (Fig. 5B) is plausible, despite the mod-
management factors play a crucial role in this respect. est R2 values of the fitted curve. Richards and Mendez
Consistent with this, Birhane et al. (2020) attributed (2014), however, argued that since AFS represents
the greater C stocks in the high elevation zone to the human-managed systems with substantial disturbance,
higher frequency of the bigger diameter class individ- greater species diversity may not be associated with
uals, owing to the exploitation of trees in the HGs at increased standing stock of C, implying the crucial
the mid and lower elevations. Furthermore, in the role of managerial interventions (species introduction,
tropical mountain forests of Colombia, Phillips et al. planting, harvesting, and other agronomic practices) in
(2019) reported a unimodal distribution of above- determining biomass C stocks. A more important
ground C stocks with elevation peaking in the middle driver of C sequestration than species richness,
of the gradient, implying a more complex interrela- therefore, is the number of stems per garden or
tionship between elevation and biomass C stocks. stocking levels (Fig. 5A).
As documented by Kumar (2011) for the mid and With a total of 753 plant species (313 herbs, 155
lower elevation zones of central Kerala, the size of shrubs, and 285 trees; Table 2), the species richness of
holdings may systematically vary the C stocks of the the HGs in the study area was profound. Not only that,
HG systems. However, the influence of holding size on but the Kerala HGs also harbored many species that
the biological C sequestration process was modest in figure in the IUCN Red List. For instance, a total of 43
the present study (Fig. 3B). Kumar (2011) in the study Threatened, Endangered, Critically Endangered, and
on the aboveground C stocks of HGs for a subset of the Vulnerable plants were located in the study area, of
sampled locations reported that C stocks (on a unit which 14 were endemic to the Western Ghats
area basis) decreased in the order, small [ medium [ (Table 3). Incidentally, Kerala is home to 5094 taxa
large size classes. Saha et al. (2009) also reported of flowering plants (Sasidharan 2012), which includes
similar results for soil organic C stocks in the HGs of 1709 taxa that are endemic to peninsular India (Nayar
Thrissur, Kerala. The inclusion of a wider elevational et al. 2008). In addition, the HGs contained several
belt for sampling in the present study (Table 1) and the exotic plants as well (data not presented). According
disparate array of HG sizes, besides the interplay of to Sasidharan (2012), the flowering plants of Kerala
many site factors and managerial interventions of include 858 exotics; of which around 200 species have
variable intensity, may explain this weak relationship become naturalized. Furthermore, the farmers may
(Fig. 3B). The bottom line is that site factors, focus on exotic or other fast-growing timber tree
123
Agroforest Syst
species because of economic benefits (Nath et al. species diversity over time (Dawson et al. 2013; Nair
2016) or promotion by certain external agencies et al. 2021).
(Kabir and Webb 2008b). Consistent with this, Nath Although harbingers of biodiversity and aligned
et al. (2016), evaluating the tree planting behavior of with SDG-15, HGs have received little or no attention
farmers of Kodagu district, Karnataka state (India), from conservationists in the past. Indeed, there are no
reported that exotic species such as Grevillea robusta policy prescriptions to promote HGs in Kerala and
are planted 5.4 times more often than native trees. elsewhere. Clearly, there is a need for incentivizing the
The Simpson’s and Shannon’s diversity indexes, farmers to conserve these biodiverse land use systems
the two popular diversity indexes used in this study, by repaying the cost of maintaining native trees on
for the arboreal species generally followed a similar farms, as exemplified in the Payment for Ecosystem
trend (Table 4). The diversity indexes were also Services (PES) schemes (Ferraro and Kiss 2002;
moderately high. The natural forest ecosystems, Pascual et al. 2014).
however, had substantially higher diversity indexes
(unpublished data). Holding size (Fig. 6A) and eleva-
tional gradient (Fig. 6B) exerted only a modest Conclusions and way forward
negative influence on Simpson’s floristic diversity
attributes. Earlier, Kumar (2011) hypothesized that This study evaluated aboveground C stocks and
smaller gardens may have higher species richness floristic diversity attributes of HGs of varying ecolog-
(unit area basis) than larger gardens under identical ical characteristics along an elevational gradient from
ecological conditions (e.g., mid and lower-elevation the coastal plains to the High Ranges of Kerala and,
HGs). For a wide range of altitudinal positions and inter alia, examined the association between tree
holding sizes, however, the variations in site charac- stocking levels and species richness with biomass C
teristics and managerial interventions may over- stocks and garden sizes. Aboveground biomass C
shadow such effects, and as a result, the stocks of Kerala HGs showed a modest negative
generalization that small HGs may have higher species correlation with elevation. Although a robust relation-
richness than larger gardens is unlikely to hold ship was expected, this was not evident owing to the
(Fig. 6A). Likewise, the notion that farmers may interplay of a wide array of locational and manage-
grow/plant more species at lower elevations (e.g., ment factors. Indeed, garden management was highly
Birhane et al. 2020) resulting in greater species individualistic in the study area (author’s observa-
diversity at lower elevations was also not true for the tions), making any generalization on this vis a vis
HGs of central Kerala (Fig. 6B). Furthermore, the elevation, difficult. As expected, C stocks (per ha) and
high evenness indexes (Table 4) indicate that individ- holding sizes showed a negative trend, albeit weak,
uals are evenly distributed across species, with no implying greater biomass accumulation on a unit area
discernible influence from location and, inter alia, basis in the smaller gardens, presumably because of
elevation. more intense land management including higher stand
Clearly, HGs provide a potentially valuable con- density (stems per ha), than large gardens. But large
servation tool that can offset the erosion of farmland gardens, by default, had greater stocking levels (stems
species diversity. Traditionally, the concept of ‘‘Bio- per garden) and higher species richness and hence
diversity hotspots’’ has been the ‘‘mantra’’ for con- higher total C stocks (per garden) than small gardens.
servation efforts, which, however, leaves out vast High species assemblages presumably have promoted
stretches of the globe with considerable biological higher net primary productivity as they harbored
diversity ‘‘in the cold’’—the so-called biodiversity divergent species with seemingly strong resource-
‘‘coldspots’’ (Kareiva and Marvier 2003; Nair et al. utilization characteristics. Both stocking levels (stems
2021). Conservation, in the future, will need to rely per garden) and species richness (species per garden)
increasingly on these smallholder-farm circa situm showed a positive trend with biomass C stocks, but
approaches such as AFS and HGs (Bhagwat et al. holding size and elevation exerted a weak negative
2008), perhaps transitioning to lower and less valuable linear relationship on Simpson’s floristic diversity
(from a conservation perspective) farmland tree index.
123
Agroforest Syst
The generally high floristic diversity levels Scientist at Kerala Agricultural University, Thrissur. Mr.
observed in the study area, which includes many rare R. Abijith, Mr. P.S. Sreebin, and Mr. Sonik Anto assisted in
the fieldwork and Dr. T.K. Kunhamu, Professor and Head,
and endangered species (43 IUCN Red-Listed spe- Department of Silviculture and Agroforestry, Kerala
cies), imply that HGs provide a potentially valuable Agricultural University, Thrissur, provided the necessary
conservation tool, i.e., a circa situm reservoir of office and laboratory space.
biodiversity that can be employed for offsetting land-
Author contributions BMK conceived the study, collected
use pressure and improving rural livelihoods by
and analyzed the data, and prepared the manuscript.
providing a range of goods and services. The diverse
range of timber and non-timber produce-yielding Funding This research was funded under the Emeritus
species in the HGs may also alleviate the resource- Scientist Scheme of the Indian Council of Agricultural
use pressure on natural forests. The potential of tree- Research, New Delhi, which was awarded to the author as per
Letter F. No. 9 (24)/2018-ES-HRD dated 2 July 2018 of the
covered but intensively managed landscapes to sup- Indian Council of Agricultural Research, New Delhi. The author
port native species suggests that maintaining and also reports a non-financial support relationship with Kerala
creating habitats in human-dominated landscapes can Agricultural University, Vellanikkara that includes office and
help conserve a large proportion of biodiversity. administrative support.
Furthermore, conservation in the future will need to
Data availability Original data are available with the author,
rely increasingly on these smallholder-farm circa which can be shared upon request. Analyzed and summary data
situm approaches, often dubbed as biodiversity ‘‘cold- are presented in the manuscript along with an Appendix
spots’’. However, HGs in Kerala have received little or describing the names of 753 plant species.
no formal attention from the perspectives of biodiver-
sity conservation or climate change mitigation, Code availability Not applicable.
although the Sri Lankan HGs figure prominently in Declarations
that country’s climate change mitigation programs.
From the multifunctional land-use planning per- Conflict of interest The author declares that he has no known
spective, the conservation feature of HGs should be competing financial interests or personal relationships that could
have appeared to influence the work reported in this paper. The
given proper recognition. However, biodiversity con- author further declares that the research was conducted in the
servation in managed ecosystems such as HGs should absence of any commercial or financial relationships that could
be sensitive to the economic needs of the farmers as be construed as a potential conflict of interest.
well. The farmers often must sacrifice part of their
Consent for publication Not applicable.
incomes for maintaining high species diversity on the
farmlands, which should be compensated. Unless this Ethical approval An ethics approval was not required as per
is done, more and more of the traditional land use the Kerala Agricultural University guidelines and national reg-
systems are likely to be transformed into intensive ulations. However, the expert committee of the Indian Council
of Agricultural Research, New Delhi has reviewed and approved
agricultural production systems with a fewer number
the study as a competitive research grant.
of species. In fact, many multi-species Kerala HGs
have been converted into monospecific production Consent to participate Verbal informed consent was
systems of rubber and coconut primarily for greater obtained from all research participants (180 farmer respondents)
of the field study according to institutional and national
economic gains in the past. Therefore, incentive
requirements and no minors (non-adults) participated in this
schemes should be designed to repay the cost of research. Verbal approval (prior) from the head of the family
maintaining these biodiverse agroecosystems involv- also was obtained for conducting the field survey.
ing a diverse range of native trees, i.e., payment for the
ecosystem services provided by HGs, besides soft
credits for developing ecological farming systems to Appendix 1
make conservation appealing to the farmers.
See Table 5.
Acknowledgements The study described in this paper was
conducted during the tenure of the author as an ICAR Emeritus
123
Agroforest Syst
Table 5 List of herbs, shrubs, trees, and pteridophytes observed in the homegardens of central Kerala and their IUCN Redlist status
Sl. no: Species name Local name Habit IUCN – Red list category
1. Herbs
1. Abrus precatorius Kunni kuru H NE
2. Achyranthes aspera Kadaladi H NE
3. Achyranthes spp. – H NE
4. Acorus calamus Vayambu H LC
5. Acmella calva Eripacha H NE
6. Actinidia deliciosa Kiwi H NE
7. Acmella paniculata – H LC
8. Adenosma Indiana Kasithumba H LC
9. Adenostemma lavenia Sticky Daisy H NE
10. Aerva lanata Cherula H NE
11. Ageratum conyzoides Appa H LC
12. Ageratum houstonianum Neela appa H NE
13. Ageratina adenophora Neelagiri H NE
14. Aglaonema commutatum Aglaonema H NE
15. Alangium salviifolium ssp. hexapetalum Valli ankolam C NE
16. Aloe vera Kattar vazha H NE
17. Alocasia amazonica Elephant’s ear H NE
18. Alternanthera paronychioides Smooth chaff flower H NE
19. Alternanthera tenella – H NE
20. Alternanthera bettzickiana – H NE
21. Alternanthera brasiliana Joy weed H NE
22. Allium sativum Veluthulli H NE
23. Alpinia calcarata Chittaratha H NE
24. Amaranthus spinosus Mullan cheera H NE
25. Amaranthus viridis Cheera H NE
26. Amaranthus tricolor Red Amaranthus H NE
27. Amorphophallus bulbifer Kattuchena H NE
28. Amorphophallus paeoniifolius Chena H LC
29. Ananas comosus Pineapple H NE
30. Alternanthera bettzickiana – H NE
31. Anamirta cocculus Nanchuvalli C NE
32. Andrographis paniculata Kiriyathu H NE
33. Anthurium sp. Anthurium H NE
34. Anthurium andraeanum Anthurium H NE
35. Axonopus compressus Buffalo grass H NE
36. Asclepias curassavica Kammalchedi H NE
37. Axonopus spp. – H NE
38. Asystasia gangetica Chinese violet H NE
39. Asystasia intrusa Common Asystacia H NE
40. Asparagus setaceus Evergreen C NE
41. Bacopa monnieri Bhramhi H LC
42. Barleria cristata Kattu Kanakambaram H NE
43. Barleria prionitis Kanakambaram H NE
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
Table 5 continued
Sl. no: Species name Local name Habit IUCN – Red list category
123
Agroforest Syst
123
Agroforest Syst
Kumar BM, Nair PKR (2004) The enigma of tropical home- matters in payments for ecosystem services. Bioscience
gardens. Agrofor Syst 61(1):135–152. https://doi.org/10. 64(11):1027–1036. https://doi.org/10.1093/biosci/biu146
1023/B:AGFO.0000028995.13227.ca Phillips J, Ramirez S, Wayson C, Duque A (2019) Differences in
Kumar BM, Takeuchi K (2009) Agroforestry in the Western carbon stocks along an elevational gradient in tropical
Ghats of peninsular India and the Satoyama landscapes of mountain forests of Colombia. Biotropica 51(4):490–499.
Japan: a comparison of two sustainable land use systems. https://doi.org/10.1111/btp.12675
Sustain Sci 4:215–232. https://doi.org/10.1007/s11625- Raich JW, Russell AE, Kitayama K, Parton WJ, Vitousek PM
009-0086-0 (2006) Temperature influences carbon accumulation in
Leuschner C, Moser G, Bertsch C, Roderstein M, Hertel D moist tropical forests. Ecology 87:76–87. https://doi.org/
(2007) Large altitudinal increase in tree root/shoot ratio in 10.1890/05-0023
tropical mountain forests of Ecuador. Basic Appl Ecol Richards MB, Mendez VE (2014) Interactions between carbon
8:219–230. https://doi.org/10.1016/j.baae.2006.02.004 sequestration and shade tree diversity in a smallholder
Malhi Y, Girardin CAJ, Goldsmith GR, Doughty CE, Salinas N, coffee cooperative in El Salvador. Conserv Biol
Metcalfe DB, Huasco WH, Silva-Espejo JE, del Aguilla- 28:489–497. https://doi.org/10.1111/cobi.12181
Pasquell J, Amézquita FF, Aragão LEOC, Guerrieri R, Saha SK, Nair PKR, Nair VD, Kumar BM (2009) Soil carbon
Ishida FY, Bahar NHA, Farfan-Rios W, Phillips OL, Meir stock in relation to plant diversity of homegardens in
P, Silman M (2017) The variation of productivity and its Kerala, India. Agrofor Syst 76(1):53–65. https://doi.org/
allocation along a tropical elevation gradient: a whole 10.1007/s10457-009-9228-8
carbon budget perspective. New Phytol 214:1019–1032. Sasidharan N (2012) Flowering plants of Kerala—version 2.0.
https://doi.org/10.1111/nph.14189 DVD No. 14. Kerala Forest Research Institute, Peechi,
Marshall AR, Willcock S, Platts PJ, Lovett JC, Balmford A, Kerala, India
Burgess ND, Latham JE, Munishi PKT, Salter R, Shirima Selmants PC, Litton CM, Giardina CP, Asner GP (2014)
DD, Lewis SL (2012) Measuring and modelling above- Ecosystem carbon storage does not vary with mean annual
ground carbon and tree allometry along a tropical elevation temperature in Hawaiian tropical montane wet forests.
gradient. Biol Conserv 154:20–33. https://doi.org/10.1016/ Glob Change Biol 20:2927–2937. https://doi.org/10.1111/
j.biocon.2012.03.017 gcb.12636
Mattsson E, Ostwald M, Nissanka SP (2018) What is good about Shanmughavel P, Francis K (1996) Aboveground biomass
Sri Lankan homegardens with regards to food security? A production and nutrient distribution in growing bamboo
synthesis of the current scientific knowledge of a multi- (Bambusa bambos (L.) Voss). Biomass Bioenergy
functional land-use system. Agroforest Syst 92: 10(6):383–391. https://doi.org/10.1016/0961-
1469–1484. https://doi.org/10.1007/s10457-017-0093-6 9534(95)00124-7
Moser G, Leuschner C, Hertel D, Graefe S, Soethe N, Iost S Shannon CE, Weaver W (1949) The mathematical theory of
(2011) Elevation effects on the carbon budget of tropical communication. University of Illinois Press, Urbana, p 125
mountain forests (S Ecuador): the role of the belowground Sharma R, Mina U, Kumar BM (2022) Homegardens agro-
compartment. Global Change Biol 17:2211–2226. https:// forestry systems in achievement of sustainable develop-
doi.org/10.1111/j.1365-2486.2010.02367.x ment goals: a review. Agron Sustain Dev 42:44. https://doi.
Nair PKR, Kumar BM, Nair VD (2009) Agroforestry as a org/10.1007/s13593-022-00781-9
strategy for carbon sequestration. J Plant Nutr Soil Sci Shorrocks VM, Templeton JK, Iyer GC (1965) Mineral nutri-
172:10–23. https://doi.org/10.1002/jpln.200800030 tion, growth and nutrient cycle of Hevea brasiliensis: 3.
Nair PKR, Kumar BM, Nair VD (2021) An introduction to Relationship between girth and shoot dry weight. J Rubber
agroforestry: four decades of scientific developments, 2nd Res Inst Malaya 19:5–92
edn. Springer Science, The Netherlands Simpson EH (1949) Measurement of diversity. Nature
Nair PKR, Nair VD, Kumar BM, Showalter JM (2010) Carbon 163:688–1688. https://doi.org/10.1038/163688a0
sequestration in agroforestry systems. Adv Agron Sundqvist MK, Sanders NJ, Wardle DA (2013) Community and
108:237–307. https://doi.org/10.1016/S0065- ecosystem responses to elevational gradients: processes,
2113(10)08005-3 mechanisms, and insights for global change. Annu Rev
Nath CD, Schroth G, Burslem DFRP (2016) Why do farmers Ecol Evol Syst 44:261–280. https://doi.org/10.1146/
plant more exotic than native trees? A case study from the annurev-ecolsys-110512-135750
Western Ghats, India. Agric Ecosyst Environ Thompson ID, Ferreira J, Gardner T, Guariguata M, Koh LP,
230:315–328. https://doi.org/10.1016/j.agee.2016.05.013 Okab K, Pan Y, Schmitt CB, Tylianakis J, Barlow J, Kapos
Nayar TS, Sibi M, Rasiya Beegam A, Mohanan N, Rajkumar G V, Kurz WA, Parrotta JA, Spalding MD, van Vliet (2012)
(2008) Flowering plants of Kerala: status and statistics. Forest biodiversity, carbon and other ecosystem services:
Rheedea 18:95–106 relationships and impacts of deforestation and forest
Pascual U, Phelps J, Garmendia E, Brown K, Corbera E, Martin degradation (Chapter 2). In: Parrotta JA, Wildburger C,
A, Gomez-Baggethun E, Muradian R (2014) Social equity Mansourian S (eds) Understanding relationships between
123
Agroforest Syst
biodiversity, carbon, forests and people: the key to Guinea. Glob Change Biol 23(11):4873–4883. https://doi.
achieving REDD? objectives. A global assessment report org/10.1111/gcb.13741
prepared by the Global Forest Expert Panel on Biodiver-
sity, Forest Management, and REDD?. IUFRO World
Publisher’s Note Springer Nature remains neutral with
Series Volume 31. Vienna, p 161
regard to jurisdictional claims in published maps and
Tilman D, Lehman CL, Thomson T (1997) Plant diversity and
institutional affiliations.
ecosystem productivity: theoretical considerations. Proc
Natl Acad Sci USA 94:1857–1861. https://doi.org/10.
1073/pnas.94.5.1857 Springer Nature or its licensor (e.g. a society or other partner)
Venter M, Dwyer J, Dieleman W, Ramachandra A, Gillieson D, holds exclusive rights to this article under a publishing
Laurance S, Cernusak LA, Beehler B, Jensen R, Bird MI agreement with the author(s) or other rightsholder(s); author
(2017) Optimal climate for large trees at high elevations self-archiving of the accepted manuscript version of this article
drives patterns of biomass in remote forests of Papua New is solely governed by the terms of such publishing agreement
and applicable law.
123