J. For. Res.

DOI 10.1007/s11676-016-0317-z

ORIGINAL PAPER

Variation in seed traits and seedling vigour of Cordia africana

Lam. provenances in Ethiopia
Abayneh Derero1 • Genene Tesfaye1 • Zewdie Woldemariam1

Received: 29 October 2015 / Accepted: 7 March 2016

 Northeast Forestry University and Springer-Verlag Berlin Heidelberg 2017

Abstract Cordia africana is a very important indigenous Determining quantitative variations in seed traits and
tree species for timber and other products and services in seedling vigour among provenances and the patterns along
Ethiopia. Seed traits and seedling vigour of seeds from 12 environmental gradients are essential for informing deci-
provenances of the species were compared. Mean seed sions on the tree improvement programme of the species.
mass for each provenance was assessed for 1000 seeds in
10 replications, and seed length and width were measured Keywords Clinal  Correlation  Environmental gradient 
for 100 seeds in four replications per provenance. Germi- Improvement  Quantitative
nation was assessed in a glasshouse taking 400 seeds from
each provenance. Root collar diameter and height of 440
seedlings grown in a nursery for 290 days in Addis Ababa Introduction
were measured. Seed length, seed width, seed mass, and
shoot height and root collar diameter of seedlings showed Understanding variation among provenances and progenies
significant differences (P \ 0.001) among provenances. is essential in tree improvement programmes so as to
Correlation analysis between seed and seedling traits and realize genetic gains through selections (Weber et al.
environmental variables revealed significant positive cor- 2008). Studies on quantitative traits are vital for under-
relations between seed width and germination percentage, standing such genetic variation in a given species (Mamo
seed width and altitude, seed width and longitude and seed et al. 2006; Wahid et al. 2006; Rao et al. 2008; Vitasse
length and latitude. Negative correlations were obtained et al. 2009; Singh et al. 2010), and they are complementary
between seed width and temperature, seed width and to the findings that use molecular markers in mapping
rainfall, seed mass and temperature, and germination and genetic structures (e.g., Derero et al. 2011; Sertse et al.
temperature of the seed source. As expected, differences in 2011). Traits such as growth performance and adaptation
seed trait did not explain the variability in seedling vigour. are mainly quantitative in nature, and they are controlled
by polygenic traits (White et al. 2007).
Hence, it is important to understand the performance of
Project funding: This work was supported by the Ethiopian a given tree species with respect to certain traits by
Environment and Forest Research Institute as part of the Tree establishing a common garden experiment or analysing
Improvement Programme for the species. their performances in the nursery or greenhouse. Some
traits such as seed shape, seed colour, pod shape, flower
The online version is available at http://www.springerlink.com.
colour, pod colour and flower position can serve as mor-
Corresponding editor: Zhu Hong. phological genetic markers in studies of inheritance
(Vendramin and Hansen 2005). Different provenances of a
& Abayneh Derero given species may exhibit differences in their seed trait due
[email protected]
to genetic and environmental differences (Ginwal et al.
1
Ethiopian Environment and Forest Research Institute, 2005; Loha et al. 2008; Hathurusingha et al. 2010) and may
P. O. Box 30708, Addis Ababa, Ethiopia reveal clinal pattern of variation (Roy et al. 2004). Thus,

123
 A. Derero et al.

understanding variations in seed traits and seedling vigour transfers and use of the species in plantations and agro-
in various populations of a given species are essential and forestry practices. Therefore, the general objective of the
also needed for understanding physiological processes present study was to evaluate differences in seed traits and
involved in seed development and seedling growth to seedling growth among 12 provenances of C. africana in
develop sampling frameworks of populations for tree Ethiopia. The specific objectives of the study were to (1)
improvement activities and for producing seed collection determine differences in seed traits among provenances, (2)
and transfer guidelines. evaluate variability in seed germination and growth of
Seeds are plant reproductive materials and vehicles of seedlings along provenance lines, and (3) establish rela-
genes capable of exhibiting almost all of the processes tionships of biological traits within themselves and with
found in mature plants (Schmidt 2007; Copeland and environmental variables of the provenances considered for
McDonald 2001). They portray a long and complex evo- the study.
lutionary and physiological history of a given plant geno-
type, and they are conditioned to the variation of a long and
complex future (Flores 2002). They can attain different Materials and methods
sizes, colour and shape owing to genetic variability and
environmental differences. As a result, seeds from different Seed collection and processing
provenances and locations can exhibit marked differences
in their germination performances (Negash 2003; Carles Throughout Ethiopia, seeds of C. africana were collected
et al. 2009; Derero et al. 2012) including percentage, rate, from 12 sites from 15 to 26 adult trees at least 50 m apart
and time of germination (Baskin and Baskin 2001). So far, (Fig. 1) to capture the range of genetic variation. The trees
only a few studies have characterized seed multiplicity and on average had a diameter at breast height of 38.7 (±14.67)
the resulting germination and growth differences in the cm, total height of 15.3 (±4.58) m, and clear bole length of
region, and a single such study focused on our target spe- 5.9 (±2.13) m (Table 1). Ripe fruits were collected in bulk
cies, Cordia africana, but narrow geographical range and from these selected trees, and seeds were extracted at the
few provenances were compared (Loha et al. 2006). collection sites, then transported to the laboratory for fur-
Cordia africana is an indigenous tree species in Ethiopia ther drying to attain the required moisture content.
that grows in areas with altitudes between 550 and 2600 m
a.s.l. and an annual rainfall of 700–2000 mm (Friis 1992). Seed mass, moisture content and seed size
It begins flowering when the tree is 4–6 years old or even determination
younger (Negash 1995). It is a monoecious species with
complete flowers and is known to be bee-pollinated (Warfa Only properly air-dried, healthy-looking seeds were con-
1988). After pollination by insects, fruit development takes sidered for the study. Seed mass was weighed on a sensi-
almost 6 months. The fruits are edible, and seeds are dis- tive balance for 1000 seeds from each provenance arranged
persed by mammals and birds. In Ethiopia, it is commer-
cially important for its timber (Abebe and Holm 2003). The
contribution of scattered trees of C. africana to various soil
properties and its importance as a coffee shade tree in
traditional agroforestry systems has been documented
(Teketay and Tegineh 1991; Yadessa et al. 2001). The
leaves are shed heavily in the dry season serving as mulch
and contributing to nutrient cycling through decomposition
(Negash 1995; Teklay and Malmer 2004). The species is
also excellent honeybee forage since it supplies abundant
pollen and copious nectar (Fichtl and Admasu 1994).
Although seed size and seed mass has been hypothesized to
greatly vary among provenances because of genetic and
environment interactions, this variation on seed traits
cannot explain differences in seedling performance. Seeds
obtained from different provenances and grown under the
same environment were also expected to differ in growth
Fig. 1 Approximate location of the sampled provenances of Cordia
along provenance lines. Understanding such differences on
africana in Ethiopia (A Zeghie, B Kemisse, C Finoteselam, D Harar,
seed size, germination and seedling growth along prove- E Hirna, F Didessa, G Butajira, H Sokoru, I Meti, J Wondogenet,
nance lines of C. africana is instrumental in guiding seed K Boditi, L Guraferda)

123
Variation in seed traits and seedling vigour of Cordia africana Lam…

Table 1 Provenances of seeds of Cordia africana collected across Ethiopia, characteristics of trees sampled and locations
Provenance No. of Growth characters Average Latitude Longitude Mean Mean Mean
seed altitude (m (N) (E) annual annual annual
trees DBH (cm) Total height Clear bole a.s.l.) RF (mm) Max. T Min. T
(m) length (m) (C) (C)
Mean SD Mean SD Mean SD

Boditi 16 36.9 11 11.4 1.79 4.4 1.94 2050 6.57 37.52 1230 24.4 12.3
Butajira 15 34.2 11 13.5 2.36 4.6 0.91 2070 8.07 38.23 1202 25.9 12.3
Didessa 17 37.2 9.1 13.7 3.02 3.8 1.23 1475 8.40 36.21 1497 30.9 15.1
Finoteselam 24 29.8 5.8 12.5 2.51 6.0 1.68 1875 10.41 37.11 1327 29.3 12.7
Guraferda 15 40.1 11 19.0 2.49 6.1 2.05 1050 6.49 38.18 1586 29.8 15.1
Harar 16 29.4 7.3 14.4 2.35 5.7 1.14 1900 9.18 42.07 810 25.3 13.1
Hirna 16 31.4 7.9 15.2 2.79 5.3 1.36 1850 9.12 41.05 1018 27.0 11.0
Kemisse 15 36.1 8.7 8.7 2.40 5.2 1.23 1500 10.43 39.52 1071 26.1 13.2
Meti 25 61.8 15 22.8 3.10 7.6 2.60 1250 7.17 35.14 1586 29.8 15.1
Sekoru 18 39.7 8.9 18.4 3.04 7.2 2.47 1900 7.55 37.25 1333 26.1 13.1
Wondogenet 17 43.0 17 15.8 2.65 6.7 1.75 1800 7.06 38.37 1182 27.0 12.5
Zeghie 26 30.7 9.2 14.5 0.71 6.7 2.20 1850 11.41 37.19 1585 28.0 10.8
Total 219 38.7 15 15.3 4.58 5.9 2.13
RF rainfall, Max. maximum, Min. minimum, SD standard deviation of the mean

in 10 replications of 100 seeds each surpassing ISTA Collar diameter and height were measured on 40 seedlings
requirement, which necessitates measuring 100 seeds in 8 per provenance with a microcaliper and ruler, respectively,
replications. Then the 1000-seed mass of each provenance 290 days after sowing.
was computed as the sum of the mass of the 10 replica-
tions. In addition, moisture content of seeds was deter- Environmental variables
mined: after taking four replications of about 5 g each of
fresh seeds, the seeds were put in oven for 2 h at 130 C, Altitude for each provenance was determined using a GPS.
and allowed to cool in desiccators with the help of silica- Climatic data (monthly rainfall, minimum and maximum
gel for 45 min. Then, oven dry mass of seeds was mea- temperature) for the majority of the sites were obtained
sured using an analytical balance, and finally the moisture from the National Meteorological Agency, and missing
content of the seeds was calculated as the ratio of mass loss data were tolerated only for 1 month per year for all the
to the fresh mass. Furthermore, seed length and seed width sites but Kemisse, which had no year with valid data on
were measured using a microcaliper from randomly temperature and hence more than one set of missing data
selected seeds of the 11 provenances; 100 seeds from each was tolerated and whatever was available was considered
provenance were arranged with 4 replications each con- for the analysis. No meteorological station was available at
taining 25 seeds. Guraferda; hence, data from the nearest station Tepi data
were used.
Assessment of germination and seedling vigour
Data analysis
The germination capacity of the seeds of the provenances
was assessed for 400 seeds each in a glasshouse for The data on seed length, seed width, seed moisture content,
43 days. Seeds of C. africana from 11 provenances were number of seeds per kilogram, 1000-seed mass, germina-
grown at the nursery of the Forestry Research Center, tion, shoot height, collar diameter, rainfall, temperature and
located at 920 N, 30430 E and 2380 m a.s.l. in Addis altitude were entered into the program SPSS version 20
Ababa; mean annual temperature maximum is 23.5 C and (IBM, Armonk, NY, USA) for analysis; percentage values
minimum is 9.8 C and mean annual rainfall is 1041 mm. were arcsine-square-root transformed to achieve normality
The seeds were sown on seed beds and transplanted into 10 for statistical tests (Steel and Torrie 1980). Of the 12
9 15 cm pots. The pots were filled with soil mix of three provenances analysed, data on Zeghie seedlings were not
parts forest soil, two parts local soil and two parts sand. generated; hence, these and other missing data were

123
 A. Derero et al.

denoted as -9 during data entry. Descriptive statistics and to 13 mm (Table 2). The mean length of seeds collected
analysis of variance were computed: univariate analysis of from Finoteselam was the highest, whereas the mean value
variance in general linear model for seed parameters and of the seeds collected from Guraferda was the smallest
one-way ANOVA for seedling traits. Mean separation was (Table 2). Statistically significant (F11,33 = 26.98,
computed using Tukey’s honestly significant difference P \ 0.001) differences were obtained among mean seed
(HSD) test. Spearman correlations were computed on the lengths from the 12 provenances considered. Pairwise
seed, seedling and environmental variables. In addition, mean comparison among the seed lengths using Tukey’s
regression coefficients (R2) were computed for mean seed HSD test revealed significant differences in most of the
width, mean 1000-seed mass, mean seed width and alti- cases.
tude, mean 1000-seed mass, mean seed width and tem- The average seed width of the 1100 seeds measured was
perature, mean seed width and rainfall as well as mean seed 6.3 (±0.91) mm, and width of individual seeds ranged from
width and longitude. In addition, the regression analysis 4 to 10 mm. The mean seed width was highest in Butajira
was conducted for individual level seed length and seed and lowest in Guraferda (Table 2). The seed width varied
width and shoot height and collar diameter. significantly (F11,33 = 43.56, P \ 0.001) among the
provenances. Pairwise mean comparison among the seed
masses of the provenances using Tukey’s HSD test
Results revealed significant differences (P \ 0.05) in most of the
cases.
Seed traits
Germination and seedling vigour
The moisture content of the air-dried seeds of the prove-
nances varied from 7.2 % for Zeghie to 11.3 % for Boditi The mean germination of C. africana in a glasshouse in
and Butajira with a mean value of 10.3 (±1.9) %. The Addis Ababa was 40 (±18.5) % and ranged from 16 % for
mean 1000-seed mass of all the 12 provenances was Guraferda and Sekoru to 69 % for Hirna (Table 3). The
231.3 g. Values for each 1000-seed varied from 160 to growth data measured for 440 seedlings indicated that the
277 g (Table 2). The mean 1000-seed mass of the 12 mean height and the mean collar diameter of the seedlings
provenances varied from 127.4 g for Guraferda to 314.7 g at the age of 290 days in the FRC nursery was 19.3 cm and
for Wondogenet provenances. ANOVA revealed signifi- 5.7 mm, respectively, and the individual measurements
cant differences among the provenances (F11,99 = 237.1, varied from 3.5 to 49.5 cm for height and from 1 to 12 mm
P \ 0.001), and the post hoc analysis with Tukey’s HSD for collar diameter. The highest mean height was achieved
test revealed significant differences (P \ 0.05) between for Finoteselam, whereas the lowest mean value was
most of the pairwise comparisons (Table 2). attained for Sekoru provenances. On the other hand, the
The average seed length of the 1200 seeds measured was highest mean collar diameter was obtained for Finoteselam
9.1 (±1.1) mm, and individual seed lengths varied from 6 and the lowest for Meti provenances (Table 3).

Table 2 Seed mass, seed

Provenance Seed mass (g) Seed length (mm) Seed width (mm)
length and seed width of seeds
of Cordia africana from 12 Range Mean SD Range Mean SD Range Mean SD
provenances
Boditi 219–225 222a 4.13 8.3–8.7 8.5a 0.90 6.1–6.4 6.2bc 0.66
b b
Butajira 253–273 263 13.91 9.5–9.8 9.6 0.75 6.7–7.0 6.9a 0.73
Didessa 190–204 197c 9.68 8.9–9.2 9.0c 0.73 6.1–6.3 6.2c 0.55
Finoteselam 265–277 271bd 8.08 9.9–10.3 10.1d 1.09 6.3–6.7 6.5a 0.89
Guraferda 124–130 127e 4.29 7.5–8.0 7.8e 1.11 4.8–5.0 4.9d 0.62
f b a
Harar 234–254 244 14.15 9.6–9.8 9.7 0.58 6.7–7.1 6.9 0.98
Hirna 239–257 248bf 12.20 8.8–9.1 9.0c 0.74 6.6–6.9 6.8a 0.71
Kemisse 247–251 249bf 3.37 9.7–10.2 10.0bd 1.19 6.2–6.5 6.3ac 0.66
Meti 160–175 168g 9.97 9.1–9.4 9.3bc 0.80 5.6–5.8 5.7b 0.62
Sekoru 184–197 191c 8.67 8.5–8.8 8.7ac 0.88 5.9–6.1 6.0b 0.63
h ac a
Wondogenet 305–324 315 13.00 8.8–9.1 9.0 0.95 6.5–6.7 6.6 0.57
Zeghie 271–293 282d 15.64 8.9–9.2 9.1ac 0.73 6.5–6.9 6.7a 0.94
Values with different letters in a column differed significantly (P \ 0.05)

123
Variation in seed traits and seedling vigour of Cordia africana Lam…

Table 3 Germination
Nos. Provenance Germination percentages Shoot height (cm) Collar diameter (mm)
percentage and mean seedling
shoot height and collar diameter Mean SE Mean SE
of seedlings of Cordia africana
from 12 provenances in Addis 1 Boditi 56 19.5 0.97 5.6 0.27
Ababa 2 Butajira 54 17.2 0.63 5.6 0.22
3 Didessa 47 23.4 1.33 5.3 0.27
4 Finoteselam 20 28.9 1.66 7.4 0.33
5 Guraferda 16 22.0 0.85 6.5 0.32
6 Harar 53 18.9 0.99 5.8 0.35
7 Hirna 69 18.5 1.10 6.1 0.22
8 Kemisse 22 19.2 0.76 5.6 0.28
9 Meti 41 14.1 0.51 4.5 0.21
10 Sekoru 16 12.9 0.43 4.9 0.27
11 Wondogenet 47 17.4 0.84 5.6 0.25
12 Zeghie NA NA NA NA NA
SE standard error of the mean, NA data not available

ANOVA revealed significant differences significant correlations were revealed between seed width
(F10,429 = 20.13, P \ 0.001) among the 11 provenances and seed mass (R2 = 0.79, P = 0.001), seed width and
grown in the nursery. Pairwise mean comparison among seed length (R2 = 0.19, P = 0.001) and shoot height and
the seedling shoot heights of the provenances using root collar diameter (R2 = 0.45, P = 0.001).
Tukey’s HSD test revealed significant differences
(F10,429 = 7.97, P \ 0.05) in 36 of the 55 comparisons,
and similar comparison among collar diameters revealed Discussion
significant differences (P \ 0.05) in 27 of the cases.
The study covered seeds of C. africana with 12 prove-
Relationships between seed traits, germination, nances and related the findings on seed and seedling traits
seedling vigour and environmental variables with each other and with environmental variables, covering
wider geographical range and much higher number of
The analysis of the Spearman correlation between seed provenances than in an earlier study on the species by Loha
traits, seedling quality and environmental variables et al. (2006). The significant difference identified in length,
revealed significant, positive correlations between mean width and mass of seeds among provenances confirm that
seed width and germination, mean seed width and altitude, provenances are important genetic entities within a species
mean seed width and longitude, and mean seed length and (e.g., Weber et al. 2008; Hathurusingha et al. 2010; Singh
latitude. In addition, significant, negative correlations were and Thapliyal 2012; Shu et al. 2012). The differentiation of
identified between mean seed width and mean minimum provenances could be due to genetic factors shaped by
temperature and mean seed width and rainfall. Further- evolutionary forces (Wright 1931; Finkeldey and Hattemer
more, the mean seed mass and the mean minimum tem- 2007) and environmental variability including geographic
perature, and the germination and the mean minimum location, altitude, climate and soils in their growing regions
temperature of the seed source were negatively correlated. (Moleele et al. 2005; Loha et al. 2006; Singh et al. 2010;
No correlation was revealed between seed traits and mean Shu et al. 2012).
maximum temperature, seed trait and seedling size, seed- The altitude of the nursery (i.e., 2380 m a.s.l.) was
ling size and all environmental variables of the seed within the species growing area range but outside of the
sources (Table 4). range of the altitude of the tested provenances, which was
The clinal pattern of seed traits along environmental generally below 2100 m a.s.l., hence, the cooler tempera-
gradients was the strongest between seed width and altitude ture at the nursery and other unforeseen factors might have
(R2 = 0.66, P = 0.001; Fig. 2). On the other hand, though retarded the growth of the seedlings. Nonetheless, the
the Spearman correlation between seed width and longi- significant difference observed among provenances in
tude was significant (Table 4), the regression analysis seedling growth (shoot height and root collar diameter) in
revealed that the variability in seed width was not the nursery receiving the same management is due to
explained by the longitudinal gradient (Fig. 2). In addition, genetic differences among the provenances, and such data

123
123
Table 4 Spearman correlation among seed traits, germination and seedling sizes of Cordia africana and environmental variables (N = 12)
Items Seed Seed 1000-Seed Shoot RCD Germination Altitude Maximum Minimum Annual Latitude Longitude
width length mass height temp. temp. rainfall

Seed width 1
Seed length 0.425 1
1000-Seed 0.742** 0.479 1
mass
Shoot height -0.041 0.100 0.055 1
RCD 0.342 0.119 0.340 0.609* 1
Germination 0.655* 0.023 0.324 -0.046 0.044 1
Altitude 0.603* 0.117 0.333 -0.191 0.056 0.460 1
Maximum -0.535 -0.053 -0.228 0.252 -0.042 -0.434 -0.804** 1
temp.
Minimum temp. -0.680* 0.064 -0.682* 0.161 -0.316 -0.674* -0.635* 0.474 1
Annual RF -0.761** -0.349 -0.459 0.077 -0.331 -0.579 -0.478 0.683* 0.382 1
Latitude 0.418 0.662* 0.524 0.191 0.223 0.055 0.133 -0.021 -0.283 -0.312 1
Longitude 0.605* 0.049 0.266 -0.082 0.461 0.411 0.267 -0.622* -0.293 -0.841** 0.098 1
RCD root collar diameter, RF rainfall, temp. temperature
** Correlation is significant at the 0.01 level, * correlation is significant at the 0.05 level
A. Derero et al.
Variation in seed traits and seedling vigour of Cordia africana Lam…

Fig. 2 Clinal pattern of seed

traits of Cordia africana along
environmental gradients as
shown by regression curves

are well supported by earlier reports (Ginwal et al. 2005; are important selection forces shaping the genetic identity
Loha et al. 2006; Singh et al. 2010). of populations since spatial variations in seed traits among
The significant correlations between the explored seed provenances can be due to differences in environmental
traits and environmental variables indicates that the latter variables (Moleele et al. 2005; Shu et al. 2012; Sáenz-

123
 A. Derero et al.

Romero et al. 2006). Clinal patterns related to some cone, plants prefer warmer climatic conditions. Common-garden
seed and seedling traits was also reported along gradients experiments need to be conducted in multiple locations
of altitude, rainfall and temperature (Roy et al. 2004; Loha along environmental gradients to reveal gene by environ-
et al. 2006; Mamo et al. 2006; Wahid et al. 2006; Shu et al. ment interactions, and hence, for selecting seeds that are
2012), and common-garden experiments would be benefi- best for local conditions in tree improvement programs.
cial if they follow such environmental gradients in realis-
tically assessing gene by environment interactions. Such Acknowledgments We thank the whole Tree Improvement Team of
the Ethiopian Environment and Forest Research Institute for all their
experiments are highly needed to differentiate genetic and support. We are also greatly indebted to the three anonymous
nongenetic inheritance as exemplified for seed size heri- reviewers for their critique and inputs. This work was supported by
tability studies by Zas and Sampedro (2015) and Singh and the Ethiopian Environment and Forest Research Institute as part of the
Sofi (2012). tree improvement programme for the species.
Furthermore, the significant correlations obtained
between seed length and width, seed width and mass, and
shoot height and collar diameter were expected, but the
References
former correlation was much weaker (R2 = 0.19) than for Abebe T, Holm S (2003) Estimation of wood residues from small-
the latter two (R2 = 0.79 and 0.45, respectively), depicting scale commercial selective logging and sawmilling in tropical
that the relationship between seed length and seed width is rain forests of south-western Ethiopia. Int For Rev 5(1):45–52
not straight forward due to high variation. Baskin CC, Baskin JM (2001) Seeds: ecology, biogeography, and
evolution of dormancy and germination. Academic, San Diego
On the other hand, the significant, negative correlation Carles S, Lamhamedi MS, Beaulieu J, Stowe DC, Colas F, Margolis
of germination with the mean minimum temperature of the HA (2009) Genetic variation in seed size and germination
seed source agrees with the data of Chunyang and Tuomela patterns and their effect on white spruce seedling characteristics.
(1997) who reported higher germination for low-tempera- Silvae Genet 58(4):152–161
Chaisurisri K, Edwards DGW, El-Kassaby AY (1992) Genetic control
ture provenances. The fact that germination percentage was of seed size and germination in Sitka spruce. Silvae Genet
only related with seed width but not with seed length and 41:348–355
seed mass may be due to a compromised seed quality due Chunyang L, Tuomela K (1997) Differentiation of seed germination
to undetected insect infestation of seeds that might have and early seedling growth in ten provenances of Eucalyptus
microtheca. J For Res 8(1):13–19
undermined germination performance of the species (Ti- Copeland LO, McDonald MB (2001) Seed science and technology,
gabu and Oden 2002). The result may also show that ger- 4th edn. Kluwer Academic Publishers, Norwell
mination is a complex physiological response and that the Derero A, Gailing O, Finkeldey R (2011) Maintenance of genetic
underlying causes for variability in germination can be diversity in Cordia africana Lam., a declining forest tree species
in Ethiopia. Tree Genet Genomes 7:1–9
multiple factors such as seed source, parental nutrition, Derero A, Eshete G, Fekadu M, Tesfaye G, Gebre B, Adele N, Abiy
seed maturity, tree age, environmental conditions during M (2012) Germination performances of provenances of Hagenia
seed development and germination environment (Chaisur- abyssinica. In: Tadesse W, Desalegn G, Yirgu A (eds) Forestry
isri et al. 1992). Our findings indicate that short-term and forest products in Ethiopia: technologies and issues.
Ethiopian Institute of Agricultural Research (EIAR), Addis
seedling growth depends on the interaction between seed Ababa, pp 24–32
source and the environment, which includes management. Fichtl R, Admasu A (1994) Honeybee flora of Ethiopia. Magraf
Verlag, Weikersheim
Finkeldey R, Hattemer HH (2007) Tropical forest genetics. Springer,
Berlin
Conclusion Flores EM (2002) Tropical tree seed biology. In: Vozzo JA (ed)
Tropical tree seed manual. Agricultural handbook no. 721. United
Results have shown marked differences among prove- States Department of Agriculture (USDA), Washington, DC
nances for various seed and seedling traits though the Friis I (1992) Forests and forest trees of northeast tropical Africa.
Kew Bulletin additional series XV. HMSO, London
pattern on germination and seedling growth differences Ginwal HS, Phartyal SS, Rawat PS, Srivastava RL (2005) Seed source
may vary across different years of seed collection and variation in morphology, germination and seedling growth of
seedling production. The study has also identified clinal Jatropha curcas Linn. in central India. Silvae Genet 54(2):76–80
differences of such traits along environmental gradients of Hathurusingha S, Ashwath N, Midmore D (2010) Provenance
variations in seed-related characters and oil content of Calo-
the origin of the seed. Therefore, sampling of populations phyllum inophyllum L. in northern Australia and Sri Lanka.
for provenance research should be framed in such a way N For 41(1):89–94
that it captures clinal differences especially taking into Loha A, Tigabu M, Teketay D, Lundkvist K, Fries A (2006)
account variables with a strong environmental gradient. In Provenance variation in seed morphometric traits, germination,
and seedling growth of Cordia africana Lam. N For 32:71–86
nurseries, the species thrives well and is widely grown in Loha A, Tigabu M, Teketay D (2008) Variability in seed- and
many places, but the life span of nursery plants varies with seedling-related traits of Millettia ferruginea, a potential agro-
agroclimatic condition and management, and generally the forestry species. N For 36:67–78

123
Variation in seed traits and seedling vigour of Cordia africana Lam…

Mamo N, Mihretu M, Fekadu M, Tigabu M, Teketay D (2006) Steel RGD, Torrie JH (1980) Principles and procedures of statistics.
Variation in seed and germination characteristics among Ju- McGraw-Hill, New York
niperus procera populations in Ethiopia. For Ecol Manag Teketay D, Tegineh A (1991) Shade trees of coffee in Hararghe,
225:320–327 eastern Ethiopia. Int Tree Crops J 7:17–27
Moleele NM, Reed MS, Motoma L, Seabe O (2005) Seed weight Teklay T, Malmer A (2004) Decomposition of leaves from two
patterns of Acacia tortilis from seven seed provenances across indigenous trees of contrasting qualities under shaded-coffee and
Botswana. Afr J Ecol 43:146–149 agricultural land uses during the dry season at Wondo Genet,
Negash L (1995) Indigenous trees of Ethiopia: biology, uses and Ethiopia. Soil Biol Biochem 36:777–786
propagation techniques. Addis Ababa University, Addis Ababa Tigabu M, Oden PC (2002) Multivariate classification of sound and
Negash L (2003) In situ fertility decline and provenance differences insect-infested seeds of a tropical multipurpose tree, Cordia
in the east African yellow wood (Podocarpus falcatus) measured africana, with near infrared reflectance spectroscopy. J Near
through in vitro seed germination. For Ecol Manag 174:127–138 Infrared Spectrosc 10:45–51
Rao GR, Korwar GR, Shanker AK, Ramakrishna YS (2008) Genetic Vendramin GC, Hansen OK (2005) Molecular markers for charac-
associations, variability and diversity in seed characters, growth, terizing diversity in forest trees. In: Geburek T, Turok J (eds)
reproductive phenology and yield in Jatropha curcas (L.) Conservation and management of forest genetic resources in
accessions. Trees 22:697–709 Europe. Arbora Publishers, Zvolen, pp 535–565
Roy SM, Thapliyal RC, Phartyal SS (2004) Seed source variation in Vitasse Y, Delzon S, Bresson CC, Michalet R, Kremer A (2009)
cone, seed and seedling characteristic across the natural distri- Altitudinal differentiation in growth and phenology among
bution of Himalayan low level pine Pinus roxburghii Sarg. populations of temperate-zone tree species growing in a common
Silvae Genet 53(3):116–123 garden. Can J For Res 39:1259–1269
Sáenz-Romero C, Guzmán-Reyna RR, Rehfeldt GE (2006) Altitudi- Wahid N, González-Martı́nez SC, Hadrami IE, Boulli A (2006) Variation
nal genetic variation among Pinus oocarpa populations in of morphological traits in natural populations of maritime pine
Michoacán, Mexico: implications for seed zoning, conservation, (Pinus pinaster Ait.) in Morocco. Ann For Sci 63:83–92
tree breeding and global warming. For Ecol Manag 229:340–350 Warfa AM (1988) Cordia (Boraginaceae) in NE tropical Africa and
Schmidt L (2007) Tropical forest seed. Springer, Berlin tropical Arabia. PhD Dissertation, Uppsala University, Uppsala,
Sertse D, Gailing O, Eliades NG, Finkeldey R (2011) Anthropogenic p 78
and natural causes influencing population genetic structure of Weber JC, Larwanou M, Abasse TA, Kalinganire A (2008) Growth
Juniperus procera Hochst. ex Endl. in the Ethiopian highlands. and survival of Prosopis africana provenances tested in Niger
Genet Resour Crop Evol 58:849–859 and related to rainfall gradients in the West African Sahel. For
Shu X, Yang X, Yang Z (2012) Variation in seed and seedling traits Ecol Manag 256:585–592
among fifteen Chinese provenances of Magnolia officinalis. Not White TL, Adams WT, Neale DB (2007) Forest genetics. CAB
Bot Horti Agrobot 40(2):274–283 International, Wallingford
Singh O, Sofi AH (2012) Variability in seed traits and genetic Wright S (1931) Evolution in Mendelian populations. Genetics
divergence in a clonal seed orchard of Dalbergia sissoo Roxb. 16:97–159
J For Res 23(1):109–114 Yadessa A, Itanna F, Olsson M (2001) Contribution of indigenous
Singh O, Thapliyal M (2012) Variation in cone and seed characters in trees to soil properties: the case of scattered trees of Cordia
blue pine (Pinus wallichiana) across natural distribution in africana Lam. in croplands of western Oromia. Ethiop J Nat
western Himalayas. J For Res 23(2):235–239 Resour 3(2):245–270
Singh B, Saklani KP, Bhatt BP (2010) Provenance variation in seed Zas R, Sampedro L (2015) Heritability of seed weight in Maritime
and seedlings attributes of Quercus glauca Thunb. in Garhwal pine, a relevant trait in the transmission of environmental
Himalaya, India. Dendrobiology 63:59–63 maternal effects. Heredity 114:116–124

123