Agroforest Syst (2021) 95:571–581

https://doi.org/10.1007/s10457-021-00609-7 (0123456789().,-volV)
( 01234567
89().,-volV)

Morphological, productive, and nutritional characterization

of Desmanthus spp. accessions under different cutting
intensities
Williane P. S. Diniz . Mércia V. F. Santos . Antonia S. C. Verás .
Márcio V. Cunha . Djalma E. Simões Neto . Toni C. Souza . José C. B. Dubeux Jr .
Daciele S. Abreu . Rinaldo L. C. Ferreira

Received: 10 October 2020 / Accepted: 23 February 2021 / Published online: 4 March 2021
Ó The Author(s), under exclusive licence to Springer Nature B.V. part of Springer Nature 2021

Abstract Rangeland legume species from the in vitro dry matter digestibility (IVDMD) were
Brazilian Caatinga play a key role as feed resources determined. There was no interaction (p C 0.05)
for livestock and are an alternative in the agroforestry between the cutting intensities and evalutaed acces-
systems. This study evaluated the effect of two cutting sions for the morphological and nutritional variables.
intensities (40 and 80 cm) on morphological, produc- The effect of accessions on stem diameter, number of
tive, and qualitative characteristics of four (5G, 6G, leaves per branch and per plant, leaf/branch ratio, and
7G, and AS) accessions of Desmanthus spp. The IVDMD was significant. The AS accession had the
experiment was established considering a completely greatest values of plant DMY and branch DMY (1.85
randomized design with subdivided plots and six and 1.27 ton DM ha-1), respectively, under the harvest
replicates at the Sugarcane Experimental Station, intensity of 40 cm. The 7G accession, however, had
Carpina/UFRPE. Every 70 days, the variables fruit- greatest DMY when harvested at 80 cm, producing
ing, pod production, number of leaves, number of 1.52 and 1.00 ton DM ha-1 for total DMY and branch
branches, leaves per branch, leaves per plant, and DMY, respectively. The Desmanthus spp. accessions
branch diameter were evaluated. The leaf/branch ratio, evaluated in this study varied in their morphological,
dry matter yield (DMY), chemical composition, and productive, and nutritional characteristics.

Keywords Cutting height  Forage  Legume 

W. P. S. Diniz  M. V. F. Santos (&)  Shrub-tree plant
A. S. C. Verás (&)  M. V. Cunha 
D. S. Abreu  R. L. C. Ferreira (&)
Federal Rural University of Pernambuco–UFRPE, Dois
Irmãos, Recife, Pernambuco 52061-160, Brazil
e-mail: [email protected] Introduction

D. E. Simões Neto With a limiting production of grains and cereals in arid

Sugarcane Experimental Station, Federal Rural University
and semiarid regions, searching for strategies to feed
of Pernambuco—UFRPE, Carpina, Pernambuco, Brazil
the animals using local resources is one of the primary
T. C. Souza targets in livestock research (Verdecia et al. 2020).
Federal Institute, Alagoinhas, Bahia, Brazil Desmanthus is a genus of legumes that occurs
throughout America and have great potential for
J. C. B. Dubeux Jr
North Florida Research & Education Center, University of improving animal production (Rangel and Gardiner
Florida, Marianna, FL, USA 2009). The forages from these plants are highly

123
572 Agroforest Syst (2021) 95:571–581

accepted by livestock and have a great seed production and, consequently, the structural characteristics of the
(Fontenele et al. 2009). plant to maintain the dynamic balance of forage
Legumes can maintain their nutritional value for a production (Silva et al. 2010).
more extended period, allowing their use in animal We hypothesized that Desmanthus spp. accessions
feeding, even during the dry season (Simbaya et al. have different morphological, DM yield, and nutri-
2020), when commonly forage grasses are already tional value and that vary with different crop man-
senescent and have low nutritional value (Kuchen- agement. Therefore, the objective of this study was to
meister et al. 2013). Furthermore, legumes may reduce evaluate the effects of different cutting intensities on
methane production during the rumen fermentation the morphological and productive characteristics and
(Vandermeulen et al. 2018). Likewise, legumes the nutritional value of Desmanthus spp. accessions
improve soil fertility and the biological activity of cultivated in the humid subtropical region of Pernam-
the soil (Cediel-Devia et al. 2019). Moreover, when buco, Brazil.
associated with diazotrophic bacteria, legumes can
carry out biological nitrogen (N) fixation (Paciullo
et al. 2017). With the potential to grow up in acidic Materials and methods
soils and with low fertility, selecting promising
rangeland legume species is of paramount importance The experiment was conducted at the Sugarcane
for feeding herds in the tropical semi-arid regions in Experimental Station of Carpina (EECAC). The
northeastern Brazil (Cook and Schultze-Kraft 2020). EECAC is a research unit of the Federal Rural
The Desmanthus genus occurs naturally in the University of Pernambuco (UFRPE), located in
Brazilian semi-arid region and it is a forage highly Carpina, North of Zona da Mata (7°510 133‘‘ S and
selected by grazing/browsing animals (Santos et al. 35°140 102’’ W). According to the Köppen classifica-
2010, 2008), and it can be ingested in natura or as hay tion, the prevailing climate in the region is As’, which
(Fontenele et al. 2009). Moreover, they are tolerant to is a rainy tropical climate with dry summers. The
harvest either by cutting and/or grazing (Kavita et al. experiment was carried out between March 2013 and
2015). Desmanthus genus naturally occurs in different January 2015. During the experimental period, the
edaphoclimatic conditions in the Brazilian semi-arid cumulative rainfall was 2206 mm, and the average
region (Costa et al. 2017) and in the world (Vander- temperature was 25 °C (Fig. 1).
meulen et al. 2018; Sonawane et al. 2019). These The soil of the experimental area is classified as
plants have desirable morphological and productive yellow Argisol, with a sandy loam texture. The 0– to
characteristics for its cultivation and use in animal 20 cm soil layer is characterized by pH 5.2, 0.28 cmolc
feeding (Calado et al. 2016). Furthermore, it can be an dm-3 of K, 0.09 cmolc dm-3 of Na, 0.1 cmolc dm-3 of
alternative for silvopastoral systems (Santos et al. Al, 5.56 cmolc dm-3 base sum, 11.16 cmolc dm-3
2019). Those legumes have high crude protein cation exchange capacity, 49.84 cmolc dm-3 base
concentration; however, most of this protein can be saturation, and 2.5% soil organic matter. The exper-
associated to cellulose fiber. In addition, these plants imental area had previously been used for the culti-
often contain tannins (Souza et al. 2020). Despite the vation of sugarcane, with mineral fertilization every
low allogamy rate reported for Desmanthus spp. year. The experimental field was established in March
(Costa et al. 2017) due to its autogamous reproduction 2013 using seedlings that had been produced from
mode, it is expected some genetic variability among seeds. It was necessary to break the dormancy of seeds
the accessions. using p.a. sulfuric acid for five minutes. Subsequently,
Management practices, such as cutting intensity, the seeds were washed with water and sown in plastic
affect the longevity, productivity, and quality of cups of 500 cm3 of substrate resulting from the
forage (Calado et al. 2016; Costa et al. 2012) due to mixture of sand, manure, and soil. The seedlings were
changes in the environment where the branches are on grown in a greenhouse with daily irrigations during
the plant. Pruning the branches may change the approximately 45 days. The seedlings were planted in
amount of radiation that affects the plant, contributing the experimental field in March 2013, measuring about
to morphological and structural adaptations. These 20 cm in height. They were transplanted into the
changes can affect the growth pattern of the branches experimental field with a spacing of 0.5 m between

123
Agroforest Syst (2021) 95:571–581 573

Fig. 1 Rainfall and average Rain Rhist Temp

air temperature from March 350 30
2013 to January 2015 at
Sugarcane Experimental 300 25

Average temperature (°C)

Station of Carpina (EECAC-
250

Rainfall (mm)
UFRPE). * Rain = rainfall; 20
Rhist = Historical rainfall 200
(dataset from 1980 to 2015); 15
Temp. = average air 150
temperature 10
100

50 5

0 0

plants and 1.0 m between rows (0.5 9 1.0 m). The with the vegetative variables (number of new leaves,
area of the experimental field was 20 9 30 m. After green leaves and senescent leaves) were scored from 0
the seedlings transplanting, periodic weeding was to 3 according to the number of leaves, fruits and pods
carried out to remove weeds. presented (0: absence; 1: until 5 leaves or fruits and
Three accessions 5G, 6G, and 7G were from the pods; 2: from 5 to 10 leaves or fruits and pods; and 3:
species Desmanthus pernambucanus (L.) Thellung more than 10 leaves or fruits and pods). The number of
(classified by the Herbarium of the Agronomic branches per plant was determined by manual count-
Institute of Pernambuco; FIB no. 02/2012). The ing. One representative branch was chosen to deter-
accessions came from seeds of ecotypes collected in mine the number of leaves per branch and branch
the municipality of occurrence—Santa Cruz do diameter from the visual analysis of the plant’s growth
Capibaribe, Pernambuco (Queiroz 2012). The fourth pattern. The leaves were counted manually, and the
accession (AS) was obtained from the Embrapa diameter was measured using a caliper. The total
Tabuleiros Costeiros in Sergipe (from the Germplasm number of leaves per plant was obtained by multiply-
Bank of Australia, Commonwealth Scientific and ing the number of leaves per branch by the total
Industrial Research Organization). number of branches of the plant. The dry matter
The accessions were evaluated under two cut production was estimated by cutting the plants at
intensities (40 or 80 cm) in a completely randomized heights of 40 or 80 cm and weighing them to obtain
study, designed in split plots, with six replicates for fresh matter. In each cut, the plants were fractionated
evaluating the morphological characteristics and five in leaf and branch, and later placed in an oven at 55 °C
replicates for determining the nutritional characteris- until they reach constant weight to determine the leaf/
tics. The accession was the major plot, and the cutting branch ratio and dry matter production.
height was the subplot. On June 2nd, 2014, standard- The chemical composition and in vitro digestibility
ized cutting was performed, in which the different of the dry matter were measured from the leaves and
cutting intensities were applied. After the initial branch (with a diameter of up to 5 mm) of each plant
cutting, three further cuttings were performed every per cut. After drying, the Desmanthus spp. accession
70-day intervals, respectively on August 11th, and samples were ground using a Wiley mill with 1-mm
October 24th, 2014, and on January 08th, 2015. The sieves, then packed in bottles for subsequent labora-
cuts were manually performed with pruning shears. torial analyses. The dry matter (DM) and protein (CP)
The morphological characteristics evaluated included concentrations were determined according to the
the number of branches per plant, number of leaves per procedure described by AOAC (2016). Neutral deter-
branch, total number of leaves per plant, and the gent fiber (NDF), acid detergent fiber (ADF), lignin,
branch diameter. and hemicellulose were analyzed according to Van
Reproductive characteristics such as fruit and pod Soest (1973). To determine the in vitro DM digestibil-
production were evaluated. These variables together ity (IVDMD), the ruminal DAISY fermenter

123
574 Agroforest Syst (2021) 95:571–581

(ANKOM Technology Corp., Fairport, NY) was used, leaves per plant. Regarding the leaf/branch ratio, the
as described by Holden (1999). 7G, 5G, and 6G did not show significant differences
Statistical tests were performed to verify whether among them and had an average value around 1.0. The
the data met the requirements for analysis of variance accession AS presented a leaf/ branch ratio equal to
(ANOVA). Data were subjected to ANOVA, and 0.62 and did not differ from 6G.
means were compared by Tukey’s test at a 5% For the total dry matter production and branch dry
significance level. An F-test was performed to com- matter production, significant interaction (p B 0.05)
pare the effects of the cutting height at a 5% was identified between cutting intensities and Des-
significance level. The following statistical model manthus spp. accessions (Table 1). However, there
was used: were no significant differences (p [ 0.05) for dry
matter production of leaves. The AS and 7G acces-
Yijk ¼ l þ Ai þ eðiÞj þ Bj þ ðABÞij þeijk
sions showed greater total DM production at cutting
where Yijk = represents the observed values in the intensities of 40 cm and 80 cm respectively. The 7G
different treatments; l = constant referring to all showed greater branch DM production at cutting
observations; Ai = effect of i-th accession; eðiÞj = plot intensity of 80 cm. On the other hand, the 5G and 6G
error; Bj = effect of the j-th cut height; ðABÞij ¼ accessions did not show statistically significant dif-
interaction between accessions and cut height; ferences between cutting heights (Table 1).
sbc = subplot error. Analyses were performed using For chemical composition variables, there was a
the SAS software (version 9.0). Correlation analyses significant effect (p B 0.05) only for the dry matter
were performed for the morphological, production, (DM) concentration of branches among the Desman-
and nutritional variables using GENES software. thus spp. accessions (Table 2). The 7G accession
presented the greatest branch DM concentration
(585 g kg-1), differing significantly from AS
Results (467 g kg-1) and 5G (461 g kg-1) accessions, but
not from 6G. The accessions of Desmanthus spp.
Evaluated morphological characteristics did not show showed average CP values of 234 and 121 g kg-1 in
significant interaction (p [ 0.05) between the cutting leaves and branches, respectively. NDF average
intensities and the Desmanthus spp. accessions. Sim- values of 373 g kg-1 DM on the leaf and
ilarly, the Desmanthus spp. accessions showed no 657 g kg-1 DM on the branch were found. ADF
significant differences for number of branches per average values of 195 and 457 g kg-1 DM on the leaf
plant, flowering, pod production, intensity of new leaf, and branch were found respectively. Regarding lignin,
green leaf, and senescent leaf (p [ 0.05), with average the accessions showed average values of 85 and
values of 12.00; 0.97; 1.55; 1.92; 2.35, and 1.75, 146 g kg-1 on leaves and branches, respectively.
respectively. Regarding to the branch diameter, num- The in vitro DM digestibility of leaf and branch of
ber of leaves per branch, number of leaves per plant Desmanthus spp. differed significantly (p B 0.05)
and leaf/branch ratio, it was observed a significant (Fig. 3). The 7G accession showed the least (p
variability (p B 0.05) among the accessions (Fig. 2). B 0.05) average IVDMD for leaves (429.5 g DM
The accession 7G had the least branch diameter kg-1) and branches (334.29 g DM kg-1). The AS and
(4.1 mm), differing significantly (p B 0.05) only with 5G accessions produced the greatest leaf IVDMD
the AS accession (6.5 mm). These accessions did not (502.8 g kg-1), followed by the 6G accession
present significant differences (p [ 0.05) with 5G and (457.7 g kg-1), which did not differ from 5G. Regard-
6G, which were similar between them. The accession ing the branch, the accessions 5G and 6G showed the
7G also had the greatest number of leaves per branch greatest IVDMD (385.3 g DM kg-1), followed by the
(15.04), with no significant differences with AS and accession AS (365.7 g DM kg-1).
total number of leaves per plant (192.5) differing only Significant positive and negative correlations of
with 6G and the total number of leaves per plant different magnitudes were found among the morpho-
(192.5). Differently, the 5G and 6G accessions shown logical, productive, and qualitative characteristics of
an average value of 10.3 leaves per branch and 123.0 the accessions (Table 3). The characteristics plant
DMY x branch DMY (r = 0.97), plant DMY x leaf

123
Agroforest Syst (2021) 95:571–581 575

Fig. 2 Morphological (a) (b) 18

variables and leaf/branch
A

Number of leaves per branch

ratio of Desmanthus spp.
0.8 16
accessions. Each histogram A 14
0.7

Branch diameter (cm)

is the mean ± standard
AB AB AB B
error (vertical bars) of 0.6 12 B
n = 12. Different capital 0.5 B 10
letters within each variable 8
0.4
indicate significant
differences (p B 0.05) 0.3 6
among accessions 0.2 4
0.1 2
0 0
AS 5G 6G 7G AS 5G 6G 7G

(c) (d)
250
A
Number of leaves per plant

A 1.2
200 AB A
1 AB

Leaf/branch ratio
150 AB 0.8 B
100 B 0.6
0.4
50
0.2
0 0
AS 5G 6G 7G AS 5G 6G 7G
Desmanthus spp. accessions Desmanthus spp. accessions

Table 1 Dry matter Variables Accessions

production (ton DM ha-1
three cuts average) of Cutting height AS 5G 6G 7G CV%
Desmanthus spp. accessions
under two cutting intensities Total DM yield 40 1.85A 0.79B 0.78B 0.74B 53.08
80 1.21AB 0.80B 0.82AB 1.52A
Level of P 0.04 0.96 0.91 0.015
Branch DM yield 40 1.27A 0.45B 0.42B 0.30B 67.64
80 0.80AB 0.38B 0.41B 0.99A
CV coefficient of variation;
ns Level of P 0.09 0.82 0.99 0.017
Not significant. Means
followed by the same Level of DM yield 40 0.41ns 0.37ns 0.58ns 0.40ns 46.81
lowercase letter within each 80 0.52ns 0.35ns 0.44ns 0.42ns
row do not differ by the Level of P 0.81 0.97 0.79 0.96
Tukey test at 5% probability

DMY (r = 0.78), leaf NDF x leaf hemicellulose Discussion

(r = 0.80), and branch ADF x branch hemicellulose
(r =—0.83) were strongly correlated with one another. The number of branches is among the factors affecting
plant development due to its potential to improve
biomass production (Silva et al. 2010). Furthermore,
the increasing number of branches may contribute to
improve the tolerance to defoliation and the regrowth
process after grazing or cut (Martuscello et al. 2015).

123
576 Agroforest Syst (2021) 95:571–581

Table 2 Chemical-bromatological composition (g DM kg-1), cutting intensities averages of different fractions of the plant,
according to the accessions of Desmanthus spp
\Variables Accessions
AS 5G 6G 7G CV %

DM leaves 424.9ns 415.2ns 439.7ns 458.4ns 18.6

DM branches 467.2 ±4.85b 460.6 ±1.24 b 538.4 ±4.71ab 585.5±0.87 a 19.2
CP leaves 229.2ns 243.2ns 227.1ns 235.1ns 6.1
ns ns ns
CP branches 124.4 122.3 120.3 117.8ns 17.5
NDF leaves 346.3ns 389.3ns 388.6ns 367.1ns 19.6
ns ns ns ns
NDF branches 659.2 652.1 650.1 666.6 5.5
ADF leaves 196.1ns 208.4ns 208.5ns 166.1ns 20.8
ADF branches 431.1ns 456.5ns 472.9ns 467.9ns 12.0
Hemicellulose leaves 164.1ns 183.5ns 157.1ns 178.8ns 25.6
Hemicellulose branches 212.8ns 202.1ns 187.5ns 205.9ns 21.7
ns ns ns ns
Lignin leaves 93.1 81.4 83.7 83.8 18.5
Lignin branches 137.8ns 142.1ns 131.7ns 137.2ns 12.3
DM dry matter, CP crude protein, NDF neutral detergent fiber, ADF acid detergent fiber.*ns Not significant. Means followed by the
same lowercase letter within each row do not differ by the Tukey test at 5% probability

Fig. 3 Leaf and branch (a) (b)

in vitro dry matter 600
A 500
Leaf IVDMD (g DM kg-1)

Branch IVDMD (g DM kg-1)

digestibility (IVDMD) (g AB
DM kg-1) of Desmanthus 500 B A
C 400 B AB
spp. accessions. Each C
histogram is the 400
mean ± standard error 300
(vertical bars) of n = 10. 300
Different capital letters 200
within each variable indicate
200
significant differences 100 100
(p B 0.05) among
accessions 0 0
AS 5G 6G 7G AS 5G 6G 7G
Desmanthus spp. accessions Desmanthus spp. accessions

There were no differences between the evaluated However, genetic factors directly affect the mor-
accessions regarding the morphological characteris- phogenic phenotypes related to various agronomic
tics, which may be associated with the plasticity of the characteristics of the plants, highlighting the potential
accessions in adapting themselves to the imposed of production and quality of the plant (Julier
management. Costa et al. (2017) when evaluating the et al.2007). The absence of significant differences in
genetic diversity of Desmanthus sp. accessions some of the morphological variables and for the
reported that the divergence among accessions was chemical composition between the accessions
mainly represented by the number of branches per (Table 2) may be the result of an autogamous
plant. reproduction mode. According to Costa et al. (2017),
Flowering, pod production, new leaf, green leaf, these plants have a low genetic variability as conse-
and senescent leaf are directly related to the nutritive quence of an allogamy rate of only 5%. Autogamous
value of the forage and persistence of the species. plants have lower intrapopulational variability when

123
Agroforest Syst (2021) 95:571–581 577

Table 3 Correlation Variables Correlation P value

between morphological,
productive, and qualitative Number leaves/branches 9 leaves CP -0.31 0.01
characteristics of
Number Leaves/Branches 9 Branches NDF 0.30 0.01
Desmanthus spp. accessions
Number leaves/branches 9 branches hemicellulose 0.41 0.05
Plant DMY 9 leaves DMY 0.78 0.05
Plant DMY 9 branches DMY 0.97 0.05
Plant DMY 9 ratio L/B -0.29 0.01
Leaves DMY 9 branches DMY 0.60 0.05
Leaves DMY 9 branches CP -0.38 0.05
Ratio L/B 9 branches IVDMD -0.42 0.05
Leaves CP 9 branches IVDMD 0.34 0.01
Branches CP 9 leaves hemicellulose 0.31 0.01
Leaves NDF 9 leaves ADF 0.70 0.05
CP crude protein, NDF
Leaves NDF 9 leaves hemicellulose 0.80 0.05
neutral detergent fiber,
DMY dry matter yield, Branches NDF 9 branches ADF 0.40 0.05
IVDMD in vitro Branches NDF 9 branches IVDMD -0.30 0.05
digestibility of dry matter, Branches ADF 9 branches hemicellulose -0.83 0.01
ADF acid detergent fiber

compared with allogamous breeding species (Brahim leaves and the leaf/branch ratio of 1.0. In addition to
et al. 2002). the productive potential, forage plants producing
The largest branch diameter presented by the AS greater amount of leaves tend to provide a better-
accessions even if it does not differ from 5 and 6G quality forage, because the leaves have the lowest fiber
(Fig. 2a) is a key factor in the production of plant concentration and the greatest concentration of nutri-
biomass. Additionally, the branch diameter can inter- ents (Muir et al. 2019).
fere with the plant’s ability to adapt to weather The AS accessions had the greatest dry matter plant
conditions and regrowth, due to its organic reserves. and leaf production (Table 1) that demonstrates its
The production and adaptation of the plant to varia- ability to persist at more intense cuttings, besides the
tions in edaphoclimatic conditions are relevant aspects probable adaptation to grazing conditions. Such access
in the identification of plants for agroforestry systems. comes from Australia; in this sense, Gardiner et al.
Desmanthus plants are naturally occurring (Costa et al. (2004) also observed adaptation of Desmanthus spp.
2017) and highly selected by grazing animals (Santos accessions under grazing conditions and severe water
et al. 2010), evidencing their potential for animal restriction in Australia. This may be due to the lower
production and agroforestry systems. apical dominance of these legumes, possibly caused
The 7G accessions shown great number of leaves by the greater elimination of meristems with the
per branch, however, it did not differ from AS cutting regime used (Silva et al. 2010). In this case, the
(Fig. 2b). This is possibly related to the greater plants when cut at 40 cm tended to emit growth points
number of growth points (meristems) along the plant in their lower parts.
after cutting. The lateral branches have a greater The greater forage production found in the AS
number of active buds responsible for the regrowth accessions at 40 cm cut intensity can also be related to
process (Silva et al. 2010), likely favoring growth and, the better light conditions provided to the remaining
consequently, the number of leaves. meristems and lower leaf/branch ratio. Greater cutting
The greater number of leaves in the branches of the intensity increases the amount of biomass removed,
plant improves the area to capture solar energy, which allowing greater light interception to the remaining
directly affects the photosynthetic process and pro- growth points. On the other hand, the greater produc-
motes greater growth and development of the forage tion observed for the 7G accessions at the 80-cm
(Calado et al. 2016). The productive potential of the cutting height (Table 1) likely resulted of increased
7G accession is highlighted by its greater number of

123
578 Agroforest Syst (2021) 95:571–581

use of the plant’s reserve carbohydrates after the the animals in response to its anti-nutritional effect and
harvesting. ability to bind with protein (Cruz et al. 2007).
The forage yield observed in the current study Plants from the Desmanthus spp. genus are consid-
differs from those reported by Calado et al. (2016) ered a supplemental source of protein for cattle.
when evaluating plants of the genus Desmanthus According to Sukkasame and Phaikaew (2011), bovi-
under cutting intensities of 20 and 40 cm in the semi- nes supplemented with fresh Desmanthus reached
arid of Pernambuco. These authors obtained an weight gains of 0.89 kg/day/animal for 90 days. Cruz
average production of 1.24 t ha-1 at cut intervals of et al. (2007) evaluated the chemical composition of
60 days. The better climatic conditions in the munic- Desmanthus virgatus and observed values of 0.5% of
ipality of Carpina possibly contributed to the increase N bound to ADF. Additionally, they reported adequate
in productivity. levels of bypass protein, and the amount of condensed
The average values of CP in the leaf and branch, tannins was enough to avoid tympanism.
and NDF and ADF for leaves (Table 2) differ from The NDF concentration in the branch fraction
those described by Medeiros et al. (2020). These (Table 2) can be characterized as a limiting factor to
authors evaluated D. pernambucanus (L.) Thellung the consumption of Desmanthus spp. by the animals.
and reported values of 240 g kg-1 and 79 g kg-1 of Plants with high NDF concentration, usually greater
CP in the leaves and branch, respectively. Likewise, than 550 g kg-1 of DM, reduce the animal feed intake
the values reported of NDF for leaves and branches and, consequently, the energy intake (Van Soest
were 336.3 g kg-1 and 521 g kg-1, respectively 1994). Hence, adequate cutting management in terms
(Medeiros et al. 2020). On the other hand, the ADF of age and height is essential to provide better quality
values reported for leaves and branch (207 g kg-1 and forage to the animals.
728 g kg-1, respectively) were greater than those The relationship between the concentration of ADF
obtained in the current study. The greater values and NDF in the branch might lead the animals to
reported by Medeiros et al. (2020) is likely due to the reduce the intake of Desmanthus spp. in response to its
irrigations performed by the authors, which provided low digestibility. According to Silva et al. (2013),
greater lengthening of the branch and synthesis of the forages with ADF values close to or greater than
cell wall. 400 g kg-1 DM have a reduced intake and digestibil-
Overall, the CP values reported here (Table 2) are ity. The assessed accessions presented variable levels
high, which shows the great biological value of the of digestibility, according to the plant fraction (Fig. 3).
plant and its potential for use in animal feed in tropical The reduced IVDMD observed for the branch fraction
regions and in silvopastoral systems (Paciullo et al. is likely associated with the proportion of thick
2014). The mean values of CP obtained, both for leaf branches in the plant and the high NDF values
and branch (233 and 121.2 g kg-1 respectively), are (Table 2), which is composed of cellulose and lignin.
greater than the minimum level of 7% recommended High levels of cellulose and lignin decrease the
by the NRC (2001). Therefore, Desmanthus spp. digestibility of the forage because part of the nutrients
accessions had enough protein to guarantee the can be bound to the fibers, especially lignin, becoming
minimum required by ruminants to stimulate volun- unavailable or barely available to the animals (Silva
tary forage intake. Furthermore, it meets the supply of et al. 2013).
N that is needed as a substrate for the bacteria Although the 7G accession (Fig. 3) has shown great
responsible for ruminal fermentation, allowing better number of leaves and the smallest branch diameters, it
use of the forage ingested. was found low levels of leaf and branch IVDMD. It is
Despite the high level of CP that is supposed to possible that this accession has constituents of difficult
meet the animal requirements during the rainy periods, digestion, such as phenolic compounds. Verdecia et al.
some native legumes may have part of these proteins (2020) reported that the presence of phenolic com-
linked to the cell wall, passing intact in the gastroin- pounds and saponins can negatively affect digestibil-
testinal tract (Santos et al. 2019). Moreover, the ity. Cruz et al. (2007) reported values of 429.0 g/kg of
proportion of condensed tannins in the plant (up to IVDMD for whole plants (branch and leaf) of
4%) may reduce the forage intake and acceptability by Desmanthus spp. during the flowering period. Queiroz
(2016) found greater values of IVDMD when

123
Agroforest Syst (2021) 95:571–581 579

evaluating the D. pernambucanus genotypes. The Altogether, it increases the production of branches and
author reported IVDMD values of 639.8 g kg-1 and dry matter of the plant as a whole.
677.86 g kg-1 for leaves and of 395.7 g kg-1 and The number of leaves/branches was positively
1426.55 g kg-1 for branches under cutting intensities correlated with NDF and branch hemicellulose. The
of 40 and 80 cm, respectively, at 75-day intervals. greater number of new and photosynthetically active
The greater values reported by Queiroz (2016) leaves stimulates the emission of new branches and
when compared with the current study is likely increases the amount of tissues with better digestibil-
associated with the difference in the month of ity, such as hemicellulose. The positive correlation
collection (March and October) and the digestibility between NDF and ADF in the leaf with the leaf
analysis methodology. The digestibility analysis car- hemicellulose concentration was already expected
ried out by Queiroz (2016) was based on washing the given that the components of the acid detergent fiber,
sample with neutral detergent instead of a pepsin mainly lignin and cellulose, are also components of
solution. The smallest value of branch IVDMD the neutral detergent fiber. Negative correlations
presented by the 7G accession (Fig. 3) is associated between branch NDF and branch IVDMD were also
with its L/B ratio (Fig. 2) that was negatively corre- reported elsewhere (Silva et al. 2013). Based on these
lated with the branch IVDMD (Table 3). findings, the NDF concentration is suggested as a
Crude protein of the leaf was positively associated pivotal factor limiting animal feed intake.
with branch IVDMD. Such association may be due to The absence of differences for chemical composi-
the greater amount of CP in the leaves (Silva et al. tion variables may be related to the fact that Desman-
2013), resulting in a greater production of carbohy- thus spp. are adapted to periods of drought (Simbaya
drates. The availability of energy from the carbohy- et al. 2020). Moreover, the presence of growth points
drates leads to the emission of new branches, which along the plant, after cutting (Martuscello et al. 2015),
have greater content of NDF and hemicellulose that may have improved the regrowth and the quality of the
have better digestibility coefficients. This assumption forage. Similarly, the soil organic matter content may
is supported by the negative relationship between the have been beneficial as well. Calado et al. (2016)
ADF and hemicellulose concentration in the branch. evaluated morphological and productive characteris-
The negative correlation between branch ADF and tics of Desmanthus spp. genotypes at two cutting
branch hemicellulose is a result of lignin and cellulose intensities (20 and 40 cm) in the semi-arid area of
being the main components of ADF (Silva et al. 2013). Brazil and reported that the genotypes were able to
These results corroborate with Zi et al. (2017) who recover after making three consecutive cuts at 60-day
also observed a positive correlation between the CP intervals.
concentration and the IVDMD of legumes. Research with a greater number of accessions, other
Regarding the productive variables, a direct rela- cutting intensities, and assessment of aspects of N
tionship was found between the dry matter production fixation is recommended. However, collecting seeds
and the production of leaf and branch fractions and maintaining these plants in cultivation conditions
(Table 3). The plant dry matter production has a direct is a limiting factor (Muir et al. 2019). It is worth
relationship with its fractions, since the production is mentioning that rangeland plants, when in cultivation,
composed by the sum of the plant fractions. The total probably present problems of pests and diseases more
dry matter production was negatively correlated with often than when cultivated in the original ecosystem,
the leaf/branch ratio. It may be due to the arboreal/ where there is natural biological control (Santos et al.
shrub behavior of these plants that increase their 2010).
production based on the greater proportion of
branches. The DM production of the leaves and the
branches were also associated as a response to the Conclusions
greater production of photosynthetically active leaves.
The increase in the number of leaves improves the Considering the conditions for cultivation in the
carbohydrates synthesis, which can be stored or used humid subtropical region of Pernambuco, the cutting
for plant maintenance and emission of new branches. intensities of 40 and 80 cm applied to the Desmanthus
spp. accessions affect the morphological

123
580 Agroforest Syst (2021) 95:571–581

characteristics and the nutritional value of the plants. gramı́neas com leguminosas sob pastejo rotacionado. Rev
The accessions showed greater CP concentration, Cien Agro-Amb 10:99–109
Cruz SESBS, Beelen PMG, Silva DS, Pereira WE, Beelen R,
highlighting its great biological value and potential for Beltrão ES (2007) Caracterização dos taninos condensados
using in animal feed in tropical regions. das espécies maniçoba (Manihot pseudoglazovii), flor de
Morphological and productive characteristics var- seda (Calotropis procera), feijão bravo (Capparis flexuosa,
ied among the Desmanthus spp. accessions. It is L) e jureminha (Desmanthus virgatus). Arq Bras Med Vet
Zootec 59:1038–1044. https://doi.org/10.1590/S0102-
suggested to cut 7G access at 80 cm, while the AS 09352007000400033
access must be cut at 40 cm to stimulate subsequent Fontenele ACF, Aragão WM, Rangel JHA, Almeida AS (2009)
growth and forage production. Leguminosas tropicais: Desmanthus virgatus (L.) Wild
The 5G and 7G accessions have morphological Uma forrageira promissora. Rev Bras Agroc 15:121–123.
https://doi.org/10.18539/cast.v15i1-4.1998
characteristics, such as branch diameter, number of Gardiner CP, Bielig A, Schlink R, Coventry WM (2004) Des-
leaves per branch and leaf/branch ratio, that are manthus- a new pasture legume for the dry tropics. Fourth
desirable for a forage plant, which reinforces their International Crop Science Congress, Brisbane
potential for cultivation. Holden LA (1999) Comparison of methods of in vitro dry matter
digestibility for ten feeds. J Dairy Sci 82:1791–1794.
https://doi.org/10.3168/jds.S0022-0302(99)75409-3
Acknowledgements The National Council for Scientific and Julier B, Huguet T, Chardon F, Chardon RA, Pierre JB, Prospero
Technological Development—CNPq, for the financial support JM, Barre P, Huyghe C (2007) Identification of quantita-
received. The Fundação de Amparo à Ciência e Tecnologia do tive trait loci influencing aerial morphogenesis in the model
Estado de Pernambuco (FACEPE), for the master’s scholarship. legume Medicago truncatula. Theor Appl Genet
This study was financed in part by the Coordenação de 114:1391–1406. https://doi.org/10.1007/s00122-007-
Aperfeiçoamento de Pessoal de Nı́vel Superior—Brasil 0525-1
(CAPES)—Finance Code 001. Kavita SR, Kumar V, Sridhar K, Vyakarnahal BS, Chanap-
pagoudar BC (2015) Effect of physical and chemical
treatments on seed dormancy and storability of hedge
References lucerne [Desmanthus virgatus (L.) Willd.]. Legume Res
38:131–136. https://doi.org/10.5958/0976-0571.2015.
00022.3
AOAC (2016) Official methods of analysis of aoac interna-
Kuchenmeister K, Kuchenmeister F, Kayser M, Wrage MN,
tional. AOAC INTERNATIONAL, Rockville
Isselstein J (2013) Influence of drought stress on nutritive
Brahim NB, Salhi A, Chtourou N, Combes D, Marrakchi M
value of perennial forage legumes. Int J Plant Prod
(2002) Isozymic polymorphism and phylogeny of 10
7:693–710
lathyrus species. Genet Resour Crop Ev 49:427–436.
Martuscello JA, Braz TGS, Silveira JM, Simeão RM, Ferreira
https://doi.org/10.1023/A:1020629829179
MR, Noronha D, Cunha FV (2015) Diversidade genética
Calado TB, Cunha MV, Teixeira VI, Santos MVFD, Cavalcanti
em acessos de Stylosanthes capitata. Bol Ind Anim 72:
HS, Lira CC (2016) Morphology and Productivity of
284–289. https://doi.org/10.17523/bia.v72n4p284
‘‘Jureminha’’ Genotypes (Desmanthus spp.) Under Dif-
Medeiros AS, Teixeira VI, Cavalcanti Filho LFM, Oliveira EN,
ferent Cutting Intensities. Rev Caatinga 29:742–752.
Silva SLF, Santos MVF (2020) Biomass production and
https://doi.org/10.1590/1983-21252016v29n326rc
chemical bromatological composition of jureminha sub-
Cediel-Devia D, Sandoval-Lozano E, Castañeda-Serrano R
mitted to increasing saline levels. Arch Zootec 69:54–64
(2019) Effects of different regrowth ages and cutting
Muir JP, Santos MVF, Cunha MV, Dubeux Júnior JCB, Lira
heights on biomass production, bromatological composi-
Júnior MA, Souza RTA, Souza TC (2019) Value of ende-
tion and in vitro digestibility of Guazuma ulmifolia foliage.
mic legumes for livestock production on Caatinga range-
Agrofor Syst 94:199–1208. https://doi.org/10.1007/
lands. Rev Bras Cien Agr 14:1–12. https://doi.org/10.5039/
s10457-019-00354-y
agraria.v14i2a5648
Cook BG, Schultze-Kraft R (2020) Clearing confusion in Sty-
NRC (National Research Council) (2001) Nutrient of require-
losanthes taxonomy: 1. S. seabrana B.L. Maass & ’t
ments of dairy cattle. National Academy Press, Washing-
Mannetje Aclarando confusiones en la taxonomı́a de Sty-
ton, DC
losanthes: S. seabrana B.L. Maass & ’t Mannetje. Trop
Paciullo DSC, Pires MFA, Aroeira LJM, Morenz MJF, Maurı́cio
Grassl-Forrajes 8:40–47. https://doi.org/10.17138/
RM, Gomide CAM, Silveira SR (2014) Sward character-
tgft(8)40-47
istics and performance of dairy cows in organic grass–
Costa JC, Fracetto GGM, Fracetto FJC, Santos MVF, Lira
legume pastures shaded by tropical trees. Animal
Júnior MA (2017) Genetic diversity of Desmanthus sp.
8:1264–1271. https://doi.org/10.1017/
accessions using ISSR markers and morphological traits.
S1751731114000767
Genet Mol Res 16:1–9. https://doi.org/10.4238/
Paciullo DSC, Pires MFA, Muller MD (2017) Opportunities and
gmr160398666
challenges of intergrated systems in animal production:
Costa RNN, Lange A, Caione G, Schoninger EL (2012) Pro-
emphasis on silvopastoral systems. Arch Latinoam Prod
dução de forragem para ovinos utilizando o consórcio de
Anim 25:1–2

123
Agroforest Syst (2021) 95:571–581 581

Queiroz IV (2016) Variabilidade genética e caracterização Van Soest PJ (1973) Collaborative study of acid-detergent fiber
morfológica, produtiva e qualitativa de acessos de Des- and lignin. J AOAC 56:781–784
manthus spp. Universidade Federal Rural de Pernambuco, Van Soest PJ (1994) Nutritional ecology of the ruminant, 2nd
Brasil, Thesis edn. Cornell University Press, Ithaca, p 476
Queiroz IV (2012) Ocorrência e germinação de sementes de Sonawane AS, Deshpande KY, Rathod SB, Shelke PR, Nikam
Desmanthus sp. coletadas no semiárido Pernambucano. MG, Gholve AU (2019) Effect of feeding Hedge lucerne
Dissertação, Universidade Federal Rural de Pernambuco, (Desmanthus virgatus) on intake, growth performance and
Brasil. body condition score in growing Osmanabadi goats. Ind J
Rangel JHA, Gardiner CP (2009) Stimulation of wool growth by Ani Sci 89:881–884
Desmanthus spp. as a supplement to a diet of Mitchell grass Souza RTA, Silva DKA, Santos MVF, Naumann HD, Magal-
hay. Trop Grass 43:106–111 hães ALR, Andrade AP (2020) Association of edaphocli-
Santos GRA, Batista AMV, Guim A, Santos MVF, Silva MJA, matic characteristics and variability of condensed tannin
Pereira VLA (2008) Determinação da composição botânica content in species from Caatinga. Rev Cienc Agron 51:1–7.
da dieta de ovinos em pastejo na Caatinga. Rev Bras Zootec https://doi.org/10.5935/1806-6690.20200042
37:1876–1883. https://doi.org/10.1590/S1516- Sukkasame P, Phaikaew C (2011) Utilization of Desmanthus
35982008001000023 virgatus as protein supplement for fattening cattle in
Santos MVF, Cunha MV, Dubeux Junior JCB, Ferreira RLC, southern Thailand. Integrated Crop-Livestock production
Lira Junior MA, Oliveira OF (2019) Native shrub-tree systems and fodder trees.
legumes of tropical America with potential for domesti- Vandermeulen S, Singh S, Ramı́rez-Restrepo CA, Kinley RD,
cation. Legume Persp 17:1–58 Gardiner CP, Holtum JA, Bindelle J (2018) In vitro
Santos MVF, Lira MA, Dubeux Júnior JCB, Guim A, Mello assessment of ruminal fermentation, digestibility and
ACL, Cunha MV (2010) Potential of Caatinga forage methane production of three species of Desmanthus for
plants in ruminant feeding. Rev Bras Zootec 39:204–215. application in northern Australian grazing systems. Crop
https://doi.org/10.1590/S1516-35982010001300023 Past Sci 69:797–807. https://doi.org/10.1071/CP17279
Silva VJ, Dubeux Júnior JCB, Teixeira VI, Santos MVF, Lira Verdecia DM, Herrera RS, Ramı́rez JL, Leonard I, Bodas R,
MA, Mello ACL (2010) Caracterı́sticas morfológicas e Andrés S, Giraldez F, Valdes C, Arceo Y, Paumier M,
produtivas de leguminosas forrageiras tropicais submetidas Santana A, Alvarez YA, Mendez Y, Lopez S (2020) Effect
a duas frequências de corte. Rev Bras Zootec 39:97–102. of age of regrowth, chemical composition and secondary
https://doi.org/10.1590/S1516-35982010000100013 metabolites on the digestibility of Leucaena leucocephala
Silva MSJ, Jobim CC, Nascimento WG, Ferreira GDG, Silva in the Cauto Valley. Cuba Agrofor Syst 94:1247–1253.
SS, Três TT (2013) Estimativa de produção e valor nutri- https://doi.org/10.1007/s10457-018-0339-y
tivo do feno de estilosantes cv. Campo Grande Semin- Zi X, Li M, Zhou H, Tang J, Cai Y (2017) Dynamics of shearing
Cienc Agrar 34:1363–1380. https://doi.org/10.5433/1679- force and its correlations with chemical compositions and
0359.2013v34n3p1363 in vitro dry matter digestibility of stylo (Stylosanthes
Simbaya J, Chibinga O, Salem AZM (2020) Nutritional guianensis) stem y Asian-Australasian. J Anim Sci
assessment of selected forages: Mulberry (Molus alba - 30:1718–1723. https://doi.org/10.5713/ajas.17.0077
Lam.), Leucaena (Leucaena luecocephala Lam de Wit.)
and Moringa (Moringa oleifera Lam.) as dry season pro-
Publisher’s Note Springer Nature remains neutral with
tein supplements for grazing animals. Agroforest Syst
regard to jurisdictional claims in published maps and
94:1189–1197. https://doi.org/10.1007/s10457-020-
institutional affiliations.
00504-7

123