Evaluating early selection in perennial tropical forages

Abstract

Perennial grass hybrids of Urochloa are evaluated for at least two years during the screening stage trials (SS) and advanced trials (AD) in breeding programs, an expensive and time-consuming process. In this study, we aimed to evaluate the potential for early selection of cultivars in this breeding scheme. We used multiple measurements of agronomic and nutritive value traits of Urochloa humidicola and Urochloa decumbens in the SS, and Urochloa ssp. in the AD. Repeatability coefficient, genetic correlation, selection efficiency (SE), and Spearman correlations were estimated. The results indicated that reliable early selection could be applied, decreasing the evaluation period to one year and a half for SS, and to one year for AD. These results were confirmed by high genetic and rank correlations, and overall SE above 50%. This proposed change in the breeding scheme could save considerable time, labor, and resources and accelerate the release of improved cultivars.

Full text

Evaluang early selecon in perennial tropical forages
291Crop Breeding and Applied Biotechnology - 19:3, 291-299, 2019
Evaluang early selecon in perennial tropical
forages
Ulisses José de Figueiredo1, Yasmin Vasques Berchembrock2,
Cacilda Borges do Valle3, Sanzio Carvalho Lima Barrios3,
Kenneth H. Quesenberry4, Patrício Ricardo Muñoz5 and José
Airton Rodrigues Nunes2*
Abstract: Perennial grass hybrids of Urochloa are evaluated for at least two
years during the screening stage trials (SS) and advanced trials (AD) in breeding
programs, an expensive and me-consuming process. In this study, we aimed to
evaluate the potenal for early selecon of culvars in this breeding scheme.
We used mulple measurements of agronomic and nutrive value traits of
Urochloa humidicola and Urochloa decumbens in the SS, and Urochloa ssp. in
the AD. Repeatability coecient, genec correlaon, selecon eciency (SE),
and Spearman correlaons were esmated. The results indicated that reliable
early selecon could be applied, decreasing the evaluaon period to one year
and a half for SS, and to one year for AD. These results were conrmed by high
genec and rank correlaons, and overall SE above 50%. This proposed change
in the breeding scheme could save considerable me, labor, and resources and
accelerate the release of improved culvars.
Keywords: Urochloa spp., repeatability coecient, forage breeding, rank cor-
relaon.
Crop Breeding and Applied Biotechnology
19:3, 291-299, 2019
Brazilian Society of Plant Breeding.
Printed in Brazil
hp://dx.doi.org/10.1590/1984-
70332019v19n3a41
ARTICLE
*Corresponding author:
E-mail: jarnunes@ua.br
ORCID: 0000-0002-6260-7890
Received: 20 November 2018
Accepted: 01 May 2019
1 Barenbrug do Brasil Sementes Ltda, 14.790-
000, Guaíra, São Paulo, SP, Brazil
2 Universidade Federal Lavras, 37.200-000,
Lavras, MG, Brazil
3 Embrapa Gado de Corte, 79.106-550, Campo
Grande, MS, Brazil
4 University of Florida, Agronomy Depart-
ment, 32611, Gainesville, Florida, United
States of America
5 University of Florida, Horcultural Science
Department, 2211 Field Hall, Gainesville,
Florida, 32611, United States of America
INTRODUCTION
The Brazilian livestock business is based on extensive grazing areas with
predominance of various Urochloa species [U. brizantha (syn. Brachiaria
brizantha), U. decumbens (syn. Brachiaria decumbens), U. ruziziensis (syn.
Brachiaria ruziziensis), and U. humidicola (syn. Brachiaria humidicola)] and
Megathyrsus maximus (syn. Panicum maximum) pastures. More than 60% of the
pasture area is covered by a few Urochloa culvars, due to their high adaptaon
to poor, acidic soils and the sasfactory performance of cale grazing these
Urochloa pastures (Jank et al. 2011, Nogueira 2012).
The Urochloa breeding program in Brazil began in decade 1980 at the
Embrapa Beef Cale Research Center in Campo Grande, Mato Grosso do Sul,
Brazil (Valle et al. 2009, Jank et al. 2011) Research rst focused on evaluaon of
a large collecon of introduced germplasm. Exploratory crosses followed, and
recurrent selecon strategies have recently been used, together with crosses of
superior apomicc plants as males with superior sexual females (Jank et al. 2014).
The early stage of the Embrapa’s forage breeding programs has produced
over 1000 individual hybrids that need to be individually evaluated under

292 Crop Breeding and Applied Biotechnology - 19:3, 291-299, 2019
UJ Figueiredo et al.
cung in Stage I trials (Screening Stage trials). In SS, the best hybrids (100-200) are evaluated for mulple traits in small
replicated plots for at least two years with mulple hand harvests per year, making it me, labor, and cost intensive.
Only 20-25 genotypes are then selected for evaluaon in Stage II (Advanced trials), which involves another two years
and two or more locaons considering the biome(s) for which the culvar will be released. In Stage II, these small plots
are harvested and evaluated for dry maer yield, regrowth, and forage nutrive value. In the last stage (Stage III or nal
trials), one to four genotypes selected from Stage II are evaluated for animal performance under grazing for another
two years in the biome(s) under consideraon. Parallel trials are carried out to screen for pest and disease resistance,
response to ferlizers, and response to abioc stresses, such as drought, water logging, toxic aluminum, etc. The forage
evaluaon process requires further grazing trials for culvar release (Alves et al. 2014, Jank et al. 2014). Altogether, the
selecon process takes between 10 to 12 years before a culvar is ready to be released.
The frequent concern of forage breeders is the number of harvests needed to select superior genotypes with a
sasfactory level of condence while conserving resources. Earlier selecon could signicantly reduce the minimum me
needed to release a culvar, but there is the concern of making a wrong decision in idencaon of superior genotypes
with fewer harvests. The repeatability coecient (ρ) represents the upper limit of heritability, and, therefore, it indicates
the eciency of predicng genotypic value from successive measurements on an individual. Thus, ρ can be used as a
parameter for determining when selecon can be made with condence (Resende 2002). Studies applying ρ for tropical
forages report 7-8 harvests are needed for plant height and dry maer weight esmates. However, 10-14 harvests for
fresh maer weight and percentage of dry maer were required to achieve coecients of determinaon greater than
80% when evaluang U. ruziziensis half-sib progenies (Souza Sobrinho et al. 2010).
The objecves of this study were to esmate repeatability coecients and genec correlaons from several Urochloa
spp. trials in order to determine the opmal number of harvests needed for selecon of genotypes for a given level of
stascal condence. The nal goal is to eventually include early selecon in the evaluaon and breeding of perennial
forages.
MATERIAL AND METHODS
Site
The eld experiments were conducted in two locaons, Campo Grande and Terenos, Mato Grosso do Sul, Brazil. In
Campo Grande, the experiment was conducted at Embrapa Gado de Corte (lat 20º 27’ 00’’ S, long 54° 37’ 00” W, alt
530 m asl), and in Terenos, the evaluaon was at the Hisaeda Farm (lat 20° 26’ 00’’ S, long 54° 51’ 00” W, alt 434 m asl).
According to the Köppen classicaon (Koek et al. 2006), the climate in both locaons is tropical rainy, subtype AW,
characterized by a well-dened dry season in winter and a rainy season in summer.
Screening stage trials (SS)
All the SS were conducted in Campo Grande. Trial 1 (T1) consisted of 50 hybrids of U. humidicola and the two
hexaploid parents. The culvar BRS Tupi was used as a male parent and the sexual ecotype as the female parent. The
experiment was set up in January 2007 from vegetave cungs in a randomized complete block design (RCBD) with eight
replicaons in 2.5 m² plots. The plots were harvested nine mes over a period of two years. Seven harvests occurred in
the rainy season (27 Nov 2007; 21 Jan, 25 Feb, 8 Oct, and 9 Dec 2008; and 28 Jan and 2 Apr 2009), and two in the dry
season (23 Apr 2008 and 12 Jul 2010).
Trial 2 (T2) consisted of 50 hybrids of U. decumbens resulng from a cross of the culvar “Basilisk” (Oram 1990) as the
male parent with three arcial tetraploid plants as sexual females. The four parents were also included as controls. The
experiment was set up as a RCBD with four replicaons, and plots of 4.0 m² were established using vegetave cungs
in December of 2010. Evaluaons began in July 2011 for six harvests: two in the dry season (20 July 2011 and 28 Sep
2011) and four in the rainy season (4 Nov 2011, 9 Dec 2011, 18 Jan 2012, and 28 Feb 2012).
A third experiment, designated Trial 3 (T3), consisted of 324 hybrids of U. decumbens laid out in an 18 x 18 simple lace
design with a plot size of 4.0 m². The harvests of this trial began in July 2012, and seven harvests were completed in one
year: 6 July 2012 and 2 Oct 2012 (dry season); and 5 Nov 2012, 11 Dec 2012, 17 Jan 2013, and 13 Mar 2013 (rainy season).

Evaluang early selecon in perennial tropical forages
293Crop Breeding and Applied Biotechnology - 19:3, 291-299, 2019
Advanced trials (AD)
For AD, only selected genotypes from SS were included in the evaluaons. Eight genotypes [2 ecotypes (U. brizantha,
B4 and B6), 2 hybrids (U. brizantha x U. ruziziensis, HBGC336 and HBGC331), and 4 commercial culvars (U. brizantha,
cv. Marandu, cv. Xaráes, and cv. BRS Piatã, and the interspecic hybrid U. brizantha x U. decumbens x U. ruziziensis, cv.
Mulato II)] were evaluated in the two locaons, Campo Grande and Terenos. In both trials, experiments were established
in a RCBD with four replicaons in January 2009. Each 15-m² plot was sown in six 4-meter rows with 0.5 m spacing
between rows. At each harvest, only the central 4.0 m² were harvested, leaving a 1.0-m border on each of the four sides.
In both locaons, plots were cut to stubble height to promote growth in April 2009, and dry maer was not recorded.
Aer that, 16 harvests were completed at each locaon over a two-year period. These experiments are idened as Trial
4 (T4) in Campo Grande and Trial 5 (T5) in Terenos, and the evaluaon of these two experiments together is Trial 6 (T6).
Traits measured
Fresh weight yield for each harvest was determined for each plot, and a subsample (300 g of fresh maer) was
used for morphological separaon into leaf, stem, and dead maer. Samples were then dried at 65 °C to calculate dry
maer percentage (DM) for each component. Then total dry maer yield (TDM, kg ha-1), leaf percentage (%L), and leaf
dry maer yield (LDM, kg ha-1) were esmated. Seven days aer harvests, plots were visually evaluated for regrowth
capacity (REG), based on the combinaon of density score (1: less than 20% of regrown llers, 2: 20% - 40%, 3: 40%
- 60%, 4: 60 - 80%, and 5: more than 80%) and rate of ller regrowth (slow, medium, and fast growth of ller height),
following the method described in Basso et al. (2009).
For forage nutrive value, dried and ground leaf samples from all trials were used, except for T3, for which forage
nutrive value data were not obtained. Crude protein (CP), neutral detergent ber (NDF), in vitro organic maer
digesbility (IVOMD), and lignin (LIG) were esmated using near infrared spectroscopy (NIRS) on a dry maer basis
(Marten et al. 1989). The NIRS was previously calibrated by comparing the results obtained in the wet chemical analyses
and the spectrum read from these same samples in the NIRS for several nutrional traits. Thus, a regres sion equaon
was esmated for each nutrional trait, using a set of samples of tropical forage grasses (Urochloa spp. and Megathyrsus
maximus) for that purpose (647 samples for CP, 613 for IVOMD, 631 for NDF, and 147 for LIG). Esmates of the coecient
of determinaon were 0.99 (CP), 0.96 (IVOMD), 0.95 (NDF), and 0.96 (LIG), showing good t of the model for predicon
of nutrional traits (data not shown).
Data analysis
Data analysis was carried out in the soware ASReml v 3.0 (Gilmour et al. 2009). For T1, T2, T3, T4, and T5, the
following mixed model was used:
y = Xm + Db + Zg + Wp + Ti + e,
where y is the data vector; m is the vector of the combined xed eects of harvest-replicaon; and b is the vector
of random sub-block eects, where b ~ NMV(0,Iσ
2
b ) and σ
2
b is the sub-block variance (this eect was considered only
for T3); g is the vector of random genotypic eects, where g ~ NMV(0,Iσ
2
g ) and σ
2
g is the genotypic variance; p is the
vector of random permanent environmental eects or plots, where p ~ NMV(0,Iσ
2
p) and σ
2
p is the variance associated
with the plot eects; i is the vector of random genotype x harvest interacon eects, where i ~ NMV(0,Iσ
2
i ) and σ
2
i is
the variance associated with the eects of the genotype x harvest interacon; and e is the vector of random errors,
where e ~ NMV(0,Iσ
2
e) and σ
2
e is the error variance. X, D, Z, W, and T are the incidence matrices of the eects m, b, g,
p, and i, respecvely.
For T6, we used the following mixed model:
y = Xm + Zg + Sp + Tgh + Qgl + Wghl + e,
where y is the data vector; m is the vector of the combined xed eect of harvest-replicaon-locaon; g is the
vector of random genotypic eect, where g ~ NMV(0,Iσ
2
g ) and σ
2
g is the genotypic variance; p is the vector of random
permanent environmental eect or plots, where p ~ NMV(0,Iσ
2
p ) and σ
2
p is the variance of plot eects; gh is the vector

294 Crop Breeding and Applied Biotechnology - 19:3, 291-299, 2019
UJ Figueiredo et al.
of the random genotype x harvest interacon eect, where gh ~ NMV(0,Iσ
2
gh ) and σ
2
gh is the variance associated with
the eects of the genotype x harvest interacon; gl is the vector of the random genotype x locaon interacon eect,
where gl ~ NMV(0,Iσ
2
gl ) and σ
2
gl is the variance associated with the eects of the genotype x locaon interacon; ghl
is the vector of the random genotype x harvest x locaon interacon eect, where ghl ~ NMV(0,Iσ
2
ghl ) and σ
2
ghl is the
variance associated with the eects of the genotype x harvest x locaon interacon; and e is the vector of random
errors, where e ~ NMV(0,Iσ
2
e ) and σ
2
e is the error variance. X, Z, S, T, Q, and W are incidence matrices of the random
eects m, g, p, gh, gl, and ghl, respecvely.
The normality assumpon of errors was checked by the normal quanle-quanle (QQ) plot, and according to the
diagnosc plot, the approximaon was adequate (Kosak and Piepho 2017). The signicance of the variance components
was veried by the likelihood rao test (LRT) (Resende 2002). The precision of the genec predicons was based on the
accuracy (r̂g̃g) computed by the following esmator: r̂g̃g = (1 – PEV /σ̂
2
g )1/2 , in which PEV is the predicon error variance
(Resende and Duarte 2007).
For each trial, the analysis of accumulated harvest was ed for the rst two harvests and then addional harvests
were added sequenally in dierent analyses unl all harvests in each trial were considered in the analysis. That way
we could dene the opmal number of harvests needed to make the selecon with high reliability. The repeatability
coecient (ρ) was esmated from the variance components by the following expressions (Falconer and Mackay 1996):
ρ = σ
2
g + σ
2
p
σ
2
g + σ
2
g + σ
2
g /rk + σ
2
e /rk
for T1, T2, T3, T4 and T5,
ρ = σ
2
g + σ
2
p
σ
2
g + σ
2
g + σ
2
gh /k + σ
2
gl /kl + σ
2
e /rkl
for T6,
where k, l, and r are the number of harvests, locaons, and replicaons, respecvely.
Addionally, the genec correlaon (rgij) between the mean of the opmal number of harvests i and the mean of all
harvests j was esmated from a bivariate mixed model for each trait using the expression rij = σgij x ( σ
2
gi x σ
2
gj )– ½, where
σ gij is the genec covariance between the mean of the opmal number of harvests i and the mean of all harvests j; σ
2
gi is the genec variance associated with the mean of the opmal number of harvests i; and σ
2
gj is the genec variance
associated with the mean of all harvests j. Whenever the standard error of the genec correlaon was at least 50%
below the esmate (stasc t =̃ 2), it was considered signicant (P<0.05). The bivariate mixed model was adjusted in a
similar way to the univariate approach as follows:
y1 = X1m1 + D1b1 + Z1g1 + e1 , for the opmal number of harvests;
y2 = X2m2 + D2b2 + Z2g2 + e2, for the mean of all harvests;
The bivariate model in matrix notaon for traits y1 and y2 can be wrien as follows:
[
y1
y2
]
=
[
X10
0X2
]
[
m1
m2
]
+
[
D10
0D2
]
[
b1
b2
]
+
[
Z10
0Z2
]
[
g1
g2
]
+
[
e1
e2
]
,
where
[
b1
b2
]
~ NMV (0,B),
[
g1
g2
]
~ NMV (0,G),
[
e1
e2
]
~ NMV (0,R),
and
B = I 
[
σ
2
b1σb12
σb12 σ
2
b2
]
, G = I 
[
σ
2
g1σg12
σg12 σ
2
g2
]
, and R = I 
[
σ
2
e1σe12
σe12 σ
2
e2
]
,
where σb12 , σg12 , and σe12 are the covariance between y1 and y2 for blocks, genotypes, and errors, respecvely; and
 is the Kronecker product.
Selecon eciency (SE) was used to check the change in ranking of the genotypes, based on the genec values,
considering the opmal number of harvests and the total number of harvests evaluated in two years. SE was esmated

Evaluang early selecon in perennial tropical forages
295Crop Breeding and Applied Biotechnology - 19:3, 291-299, 2019
using SE = A – C
B – C
x 100 (Hamblin and Zimmermann 1986), where A is the number of coincident genotypes in two
selecons; B is the number of genotypes selected [10 (T1), 4 (T4, T5, and T6)]; and C is the common number of genotypes
taken at random in two selecons (C = i x B), where i is the intensity of selecon: 20% for T1, T2, and T3 and 50% for
T4, T5, and T6.
RESULTS AND DISCUSSION
In evaluaon of plant breeding experiments, selecve accuracy is an important indicator of the reliability of selecon
since accuracy measures the correlaon between the esmates or predicons and the actual breeding values (Resende
and Duarte 2007). Thus, considering all the harvests in the SS (T1, T2, and T3), accuracy ranged from 48% (T3, CP) to
89% (T2, %L), whereas for AD, the magnitudes ranged from 69% (T6, LIG) to 97% (NDF, T4 and T6).
Genotypic variance was signicant for most traits in the trials based on the LRT (P<0.05), except for TDM (T4 and T6)
and LIG (T6). This evidence of broad genec variability allows the selecon of genotypes for agronomic and nutrive
value traits. However, the genotype x harvest interacon (GHI) eect was signicant (P<0.05) for all traits, except for
forage nutrive value traits on T1 and T6, reecng dierences in the relave performance of genotypes across harvests.
Thus, GHI directly impacts the ρ and the denion of how many harvests are necessary to adequately test a genotype.
For SS, the ρ increased as harvests were added, especially for agronomic traits, whose values were higher than for
forage nutrive value traits (Figures 1 and 2). T1 is an evaluaon of U. humidicola, and in this case, ρ increased up to a
year and a half of evaluaon, especially for agronomic traits, except for LIG. In addion, REG and CP ρ values remained
Figure 1. Repeatability coecients of the agronomic traits of
U. humidicola (Trial 1, T1) and U. decumbens (Trial 2, T2; Trial 3,
T3) from analysis of accumulated harvests (addional harvests
were added sequenally in dierent analyses unl all harvests in
each trial were considered). TDM, total dry maer yield; %L, leaf
percentage; LDM, leaf dry maer yield; REG, regrowth. Bars are
standard errors for each esmate of the repeatability coecient.
Figure 2. Repeatability coecients of the nutrive value traits of
U. humidicola (Trial 1, T1) and U. decumbens (Trial 2, T2; Trial 3,
T3) from analysis of accumulated harvests. CP, crude protein; NDF,
neutral detergent ber; IVOMD, in vitro organic maer digesbil-
ity; LIG, lignin. Bars are standard errors for each esmate of the
repeatability coecient.

296 Crop Breeding and Applied Biotechnology - 19:3, 291-299, 2019
UJ Figueiredo et al.
high aer the rst harvest. As for T2 and T3, where hybrids of U. decumbens were evaluated for one year, the ρ values
were similar to those achieved in T1. Forty-seven hybrids tested in T2 were tested in T3, and there was variability for the
traits measured in both. Genec variaon in U. decumbens was also reported by Maas et al. (2016) when there were
evaluated full-sib progenies for agronomic and nutrional value traits. Thus, the lower ρ observed in the rst harvests
for T3 was probably due to the higher phenotypic variance in T3 than T2.
In Urochloa breeding programs, normally a large number of genotypes are evaluated for two years, and several traits
are measured (Jank et al. 2014). Thus, considering our results for inial screening trials, the agronomic traits could be
reliably evaluated in a shorter me. The ρ values were above 0.80 aer six accumulated harvests in T1 and T3, and aer
four in T2 (Figure 1). According to Resende (2002), this value of ρ represents a determinaon coecient above 0.89,
which is recommended for selecon of genotypes in breeding populaons. In experiments with Megathyrsus maximus
progenies, Resende et al. (2004) reported similar results, where the ρ esmates were high, and their values increased
less than 5% for TDM and LDM aer three years of evaluaon.
The AD showed magnitudes of ρ around 0.90 for almost all traits, and above 0.90 for REG (T6) (Figure 3), CP, and
NDF (T4, T5, and T6) (Figure 4) aer the rst two harvests up to the sixteenth harvests in these trials. In contrast to the
SS, the ρ values were above 0.90 for TDM and LDM in trials T4 and T6, considering two to four cumulave harvests.
However, aer nine harvests, these esmates plateaued around 0.60, 0.70, 0.75, and 0.80 for TDM (T4 and T6) and
LDM (T4 and T6), respecvely. Besides for AD, standard errors had an overlap for accumulated harvests probably due
to the ρ values showing lile change aer the rst accumulated harvests.
Advanced trials within this forage breeding program are conducted with the goal of tesng candidate genotypes
that may ulmately become culvars. Usually, Stage II trials test less than ten genotypes, and the Brazilian Ministry of
Figure 4. Repeatability coecients of the nutrive value traits
of Urochloa ssp. in Trial 4 (T4), Trial 5 (T5), and Trial 6 (T6) from
analysis of accumulated harvests. CP, crude protein; NDF, neutral
detergent ber; IVOMD, in vitro organic maer digesbility; LIG,
lignin. Bars are standard errors for each esmate of the repeat-
ability coecient.
Figure 3. Repeatability coecients of the agronomic traits of Uro-
chloa ssp. in Trial 4 (T4), Trial 5 (T5), and Trial 6 (T6) from analysis
of accumulated harvests. TDM, total dry maer yield; %L, leaf
percentage; LDM, leaf dry maer yield; REG, regrowth. Bars are
standard errors for each esmate of the repeatability coecient.

Evaluang early selecon in perennial tropical forages
297Crop Breeding and Applied Biotechnology - 19:3, 291-299, 2019
Agriculture (MAPA) requires two years of evaluaon under cung and a further two years under grazing. Aer these
evaluaons, and considering the predicted breeding values for the agronomic and nutrive value traits of the genotypes
tested compared to commercial culvars, genotypes B6 and HBGC331 were released as new culvars. B6 was released
as the culvar BRS Paiaguás, and HBGC331 as the culvar BRS Ipyporã.
For agronomic traits, the ρ were rather variable in the early harvests but plateaued aer eight accumulated harvests
(Figure 3). TDM showed a lower ρ value compared to the other traits in T4, but TDM showed no signicant dierences
in the last harvests evaluated. For other traits, the plateau began in values above 0.80 aer two harvests, suggesng
the possibility of shorter evaluaon periods that can reduce nancial and me requirements.
MAPA also requests measurements of forage nutrive value traits. Our results indicate a situaon more favorable
than for agronomic traits, e.g., the ρ for CP and NDF were above 0.90 aer the rst harvests evaluated, and above 0.80
for IVOMD (Figure 4). Thus, the forage nutrive value traits only need to be measured for two to three harvests, mainly
for SS (Figueiredo et al. 2012). A more detailed evaluaon of forage nutrive value can be addressed through pasture
management trials prior to or aer the release of a culvar (Silva and Nascimento Júnior 2007). Nevertheless, it should
be emphasized that environmental variaon over a period of years and within years are signicant determinants of
nutrive value and forage yield (Euclides et al. 2009).
The purpose of this study was to predict an opmal number of harvests based on ρ esmates of accumulated harvests,
where if the ρ is high, mulple measures are unnecessary (Pedrozo et al. 2011). For trials T4, T5, and T6, the ρ reached
a plateau with high magnitudes aer one year of evaluaon, while for T1, this occurred in one year and a half. Thus, to
conrm if early selecon is possible, genec correlaon was esmated to determine if the level of associaon among
genotypes changed over harvests (Casler 1999). The results indicated a sasfactory correlaon between this opmal
number compared to the mean of all harvests (Table 1).
Our results suggest that with U. humidicola hybrids, the period of experimental evaluaon can be reduced by at least
a half year and they show that one year is sucient for reliable selecon of Urochloa ssp. genotypes for agronomic and
forage nutrive value traits in advanced stages. This suggeson is based on the correlaon between selecon making
based on selecon from all harvests compared to selecon from six harvests. Specically, for forage nutrive value
traits, the correlaons were almost 1.0, and the number of harvests should be even fewer than for agronomic traits.
Selecon eciency (SE) was used to evaluate the behavior of selecon considering the opmal number of harvests
based on the ρ and genec correlaon. For T1, six harvests (one year and a half) were considered, while for T4, T5,
and T6, this number was eight harvests (one year). Considering screening trials (T1) for U. humidicola, SE ranged from
62.5% (IVOMD, LIG) to 87.5% (TDM, %L, REG). Furthermore, considering ten genotypes selected from six harvests and
those selected in all nine harvests, the coincidence was high, i.e., in nine genotypes (TDM, %L, REG, NDF), eight (CP),
and seven (LDM, IVOMD, LIG) (Table 2).
Table 1. Esmates of genec correlaon and their standard deviaons (SD) between the mean of the opmal number of harvests
and the mean of all harvests for agronomic and forage nutrive value traits of U. humidicola (Trial 1, T1) and Urochloa ssp. (Trial 4,
T4; Trial 5, T5; and Trial 6, T6)
Traits
T1 T4 T5 T6
r6̅–9̅†† SD r8̅–1̅6̅††† SD r8̅–1̅6̅††† SD r8̅–1̅6̅††† SD
TDM†0.98 0.01 0.94 0.06 0.92 0.07 0.97 0.04
% L 0.95 0.02 0.95 0.04 0.87 0.09 0.91 0.07
LDM 0.99 0.01 0.96 0.04 0.93 0.06 0.96 0.04
REG 0.99 0.01 0.97 0.02 0.92 0.06 0.95 0.04
CP 0.99 0.01 1.00 0.45 0.91 0.07 0.98 0.03
NDF 0.99 0.02 0.99 0.01 0.99 0.01 0.99 0.01
IVOMD 1.00 0.02 0.98 0.03 1.00 0.01 0.99 0.02
LIG 1.00 0.06 0.94 0.05 0.74 0.23 0.88 0.12
† TDM, total dry maer yield; %L, leaf percentage; LDM, leaf dry maer yield; REG, regrowth; CP, crude protein; NDF, neutral detergent ber; IVOMD, in vitro organic
maer digesbility; LIG, lignin. †† Genec correlaon between the six (one year and a half period) and nine (two year period) harvests for U. humidicola. ††† Genec
correlaon between the eight (one year period) and sixteen (two year period) harvests for Urochloa ssp.

298 Crop Breeding and Applied Biotechnology - 19:3, 291-299, 2019
UJ Figueiredo et al.
Table 2. Number of coincident genotypes and esmates of selecon eciency (SE) (between parentheses) and Spearman correla-
on (rs) between the opmal number of harvests and all harvests evaluated (n) of trials T1, T4, T5, and T6 for agronomic and forage
nutrive value traits
Traits T1 T4 T5 T6
n (SE %) rsn (SE %) rsn (SE %) rsn (SE %) rs
TDM†9 (87.5) 0.97** 3 (50) 0.88** 3 (50) 0.95** 2 (0) 0.93**
% L 9 (87.5) 0.90** 4 (100) 0.79* 3 (50) 0.74* 2 (0) 0.71*
LDM 7 (62.5) 0.97** 4 (100) 0.76* 4 (100) 0.83** 2 (0) 0.90**
REG 9 (87.5) 0.97** 3 (50) 0.90** 3 (50) 0.83** 1 (0) 0.95**
CP 8 (75.0) 0.96** 4 (100) 0.97** 4 (100) 0.98** 1 (0) 1.00**
NDF 9 (75.0) 0.91** 3 (50) 0.98** 4 (100) 1.00** 2 (0) 0.98**
IVOMD 7 (62.5) 0.90** 2 (0) 0.78* 4 (100) 0.98** 2 (0) 0.81**
LIG 7 (62.5) 0.90** 4 (100) 0.87** 3 (50) 0.81* 3 (50) 0.67
† TDM, total dry maer yield; %L, leaf percentage; LDM, leaf dry maer yield; REG, regrowth; CP, crude protein; NDF, neutral detergent ber; IVOMD, in vitro organic maer
digesbility; LIG, lignin. *, ** Signicant by the Student t test at 5% and 1% probability, respecvely.
In the Urochloa ssp. AD, this SE was 100% for many traits in the two separate sites (T4 and T5). This conrms that one
year should be enough for selecng the best genotypes. However, when considering the combined analyses between
these two sites, the SE was 0%, with coincidence of only two genotypes of the four selected among the eight evaluated
(Table 2). This happened due to the low number of genotypes considered since the number of genotypes chosen at
random in only two selecons became biased.
The decision to reduce the me for selecon will likely create concern on the part of breeders because of lack of
condence that early selecon will be coincident with addional measurements. Thus, to conrm that early selecon
can be performed, the ranking of genotypes from the opmal number of harvests and all harvests were compared using
Spearman rank correlaons.
The Spearman rank correlaon was high for all the traits in the evaluaon trials (Table 2). Values above 0.90 were
observed for the traits in the screening trial (T1), which is evidence that selecon in one year and a half compared to
two years should fall on the same genotypes with a selecon intensity of 20%. For advanced trials, the correlaons
were above 0.71 (%L, T6), but extremely high values, above 0.95, were found, including 1.0 for NDF (T5) and CP (T6).
In conclusion, our results provide strong evidence that early selecon is possible in Urochloa breeding programs. In
screening stages that evaluate large numbers of genotypes, one year and a half is sucient for reliable selecon. Selecon
decisions are more crical in advanced trials, in which MAPA requests two years of evaluaon under harvesng and two
years for grazing evaluaon before a culvar can be registered and released for commercial use in Brazil (Jank et al. 2014).
In the advanced trials, the number of genotypes is small, and the genotypes are more stable, so our results indicated that
evaluaon could be carried out in just one year to select the best genotypes. However, addional studies in advanced
trials with more environments and genotypes should be carried out to conrm these results. Nevertheless, the results
presented here considering several trials aest that adopon of early selecon in breeding of tropical forages such as
the genus Urochloa can signicantly save me, eort, and resources without loss of reliability in releasing a culvar.
ACKNOWLEDGMENTS
The authors thank UNIPASTO, CNPq, Fundect, and Embrapa Gado de Corte for nancial support to carry out the
experiments, and CAPES for granng a doctoral degree scholarship (Finance code 001).
REFERENCES
Alves GF, Figueiredo UJ, Filho ADP, Barrios SCL and Valle CB (2014)
Breeding strategies for Brachiaria spp. To improve producvity – an
ongoing project. Tropical Grasslands-Forrajes Tropicales 2: 4-6.
Basso, KC, Resende RMS, Valle CB, Gonçalves MC and Lempp B (2009)
Avaliação de acessos de Brachiaria brizantha Stapf e esmavas de
parâmetros genécos para caracteres agronômicos. Acta Scienarum
Agronomy 31: 17-22.
Casler MD (1999) Repeated measures vs. repeated planngs in perennial
forage grass trials: an empirical analysis of precision and accuracy.
Euphyca 105: 33-2.

Evaluang early selecon in perennial tropical forages
299Crop Breeding and Applied Biotechnology - 19:3, 291-299, 2019
Euclides VPB, Macedo MCM, Valle CB, Difante GS, Barbosa RA and
Cacere ER (2009) Valor nutrivo da forragem e produção animal em
pastagens de Brachiaria brizantha. Pesquisa Agropecuária Brasileira
44: 98-106.
Falconer DS and Mackay TFC (1996) Introducon to quantave genecs.
Longman Malaysia, London, 480p.
Figueiredo UJ, Nunes JAR and do Valle CB (2012) Esmaon of genec
parameters and selecon of Brachiaria humidicola progenies using
a selecon index. Crop Breeding and Applied Biotechnology 12:
237-244.
Gilmour A, Gobel B, Cullis B and Thompsom R (2009) ASReml user guide
release 3.0. VSN Internaonal, Hemel Hempstead, 372p.
Hamblin J and Zimmermann MJO (1986) Breeding common bean for yield
in mixtures. Plant Breeding Reviews 4: 245-272.
Jank L, Barrios SC, Valle CB, Simeão RM and Alves GF (2014) The value of
improved pastures to Brazilian beef producon. Crop and Pasture
Science 65: 1132-1137.
Jank L, Valle CB and Resende RMS (2011) Breeding tropical forages. Crop
Breeding and Applied Biotechnology 11: 27-34.
Maas FI, Barrios SCL, Valle CB, Mateus RG, Marns LB and Moro GV
(2016) Esmate of genec parameters in Brachiaria decumbens
hybrids. Crop Breeding and Applied Biotechnology 16: 115-122.
Kosak M and Piepho HP (2017) What’s normal anyway? Residual plots
are more telling than significance tests when checking ANOVA
assumpons. Journal of Agronomy and Crop Science 1: 1-13.
Koek M, Grieser J, Beck C, Rudolf B and Rubel F (2006) World map of
Köppen-Geiger climate classicaon updated. Meteorologische
Zeitschri 15: 259-263.
Marten GC, Shenk JS and Barton FE (1989) Near infrared reectance
spectroscopy (NIRS): Analysis of forage quality. Agriculture handbook
(USA), n. 643, 110p.
Nogueira MP (2012) Rally da Pecuária faz diagnósco de nossas pastagens.
Balde Branco 1: 43-45.
Oram RN (1990) Register of Australian herbage plant culvars. 3rd edn.
Australian Herbage Plant Registraon Authority, Division of Plant
Industry, Common Scienc and Industry Research Organizaon
(CSIRO), Melbourne, 304p.
Pedrozo CÂ, Barbosa MHP, Silva FL, Resende MDV and Peternelli LA
(2011) Repeatability coecient of full-sib sugarcane families across
harvests and the eciency of early selecon. Euphyca 182: 423-430.
Resende MDV (2002) Genéca biométrica e estasca no melhoramento
de plantas perenes. Embrapa Informação Tecnológica, Brasília, 975p.
Resende MDV and Duarte JB (2007) Precisão e controle de qualidade
em experimentos de avaliação de culvares. Pesquisa Agropecuária
Tropical 37: 182-194.
Resende RMS, Jank L, Valle CB and Bonato ALV (2004) Biometrical analysis
and selecon of tetraploid progenies of Panicum maximum using
mixed model methods. Pesquisa Agropecuária Brasileira 39: 335-34.
Silva SC and Nascimento Júnior D (2007) Avanços na pesquisa com plantas
forrageiras tropicais em pastagens: caracteríscas morfosiológicas
e manejo do pastejo. Revista Brasileira de Zootecnia 36: 121-138.
Souza Sobrinho F, Borges V, Lédo FJS and Kopp MM (2010) Repebilidade
de caracteríscas agronômicas e número de cortes necessários para
seleção de Urochloa ruziziensis. Pesquisa Agropecuária Brasileira
45: 579-584.
Valle CB, Jank L, Resende RMS (2009) O melhoramento de forrageiras
tropicais no Brasil. Revista Ceres 56: 460-472.
This is an Open Access arcle distributed under the terms of the Creave Commons Aribuon License, which permits unrestricted
use, distribuon, and reproducon in any medium, provided the original work is properly cited.