Sélastique Doffou Akaffou · Aimé Kouassi Kouame · Nestor Bi Boh Gore ·Georges Yao Abessika · Henri Kouadio Kouassi · Perla Hamon · Sylvie Sabatier ·Jér?me Duminil

Abstract The rapid rate of deforestation in C?te d’Ivoire has led to loss of plant species diversity and also threatens some commercial tree species with extinction. Some reforestation and afforestation has been completed. However, for some species, the numbers of surviving transplants has declined. Hence, it is necessary to develop properly adapted and resilient genotypes that can effectively support the programs of forest restoration. In this context we evaluated the effects of the provenances of seeds and of various treatments on the germination rates and on plant growth of four commercial trees species (Pycnanthus angolensis, Terminalia superba, Mansonia altissima and Pterygota macrocarpa ). We analyzed these parameters for seeds of three provenances (Daloa, Akoupé and Daoukro) of four species and for thirty plants aged 6 months per provenance. Time to begin germination and time to reach maximum germination were similar among species, and were not influenced by either treatment or provenance. Untreated seeds and those soaked for 24 and 48 h in water at room temperature yielded the highest germination rates, irrespective of species or provenance. The one exception was P. angolensis, for which no germination was recorded. Within species, some provenances produced higher germination percentages than did others. No clear relationship was computed between the germination rate and the morphometric characteristics of the seeds which varied significantly between provenances.Plant growth and morphological variability also depended greatly on provenance. Genetic factors might be implicated,therefore provenance trials should be undertaken and evaluated. Furthermore, molecular analysis should be undertaken to conf irm the implications of genetic factors and allow for genetic selection.

Keywords Forest trees · Seed germination · Plant growth · Provenance · C?te d ’Ivoire

Introduction

The economic development of C?te d’Ivoire, similar to that of many developing countries, greatly depends on activities that result in forest destruction, such as agricultural expansion and forest exploitation (Verdaux and Alpha 1999).Indeed 10% of forest coverage of C?te d’Ivoire disappear each year and more than 80% of Ivorian rainforest has disappeared (FAO 2009). Another factor contributing to forest destruction is exploitation of woody trees. This has increased in recent decades when almost all parks and reserves were affected by degradation (FAO 2017). This excessive forest exploitation and agricultural expansion, plus the low rate and very slow progress of forestation and afforestation negatively impacted the biodiversity in areas already affected by the climate change (FAO 2017).

Because of recent degradation, nature conservation and forest restoration are a major concern for C?te d’Ivoire.The reconstruction of the forest area through reforestation and agroforestry are two options favored by government. Thus, in response to the Bonn challenge (2016),the country has committed to forest landscape restoration with a target of restoring five million hectares by 2030.In addition, the practice of agroforestry or planting indigenous trees species in farmlands and slogans such us zerodeforestation in cocoa (Ruf and Varlet 2017; Carodenuto 2019) are also used to encourage farmers to reforest.

In C?te d’Ivoire, Terminalia superba, Pycnanthus angolensis, Pterygota macrocarpa and Mansonia altissima are four commercial trees species highly exploited by the timber industry. These species were distributed in humid forests south of the 8th parallel. Their population sizes and stand diversities were negatively impacted by forest destruction such that among them P. macrocarpa was assessed as globally Vulnerable (Vu) for the IUCN Red List of Threatened Species (Hawthorne 1998). Afforestation, notably with T. superba and T. ivorensis has been undertaken but the trees decline after reaching 10 years of age (ITTO 2008).

Since provenance affects various traits (e.g. seed morphometrics, plant growth, resistance to biotic and abiotic stress) of tree species, one approach to improved reforestation success in C?te d’Ivoire is to develop resilient genotypes by breeding plant provenances. This was investigated for Pseudotsuga menziesii in Canada (McDermott and Robinson 1989) for resistance to disease transmitted by the pathogen Phaeocryptopus gaeumannii. In Australia, Bustos-Salazar et al. (2017) recommended forest restoration in drought and flood-prone regions using Drimys winteri var.chilensis, which is native to xeric locales. In Turkey, ?evik and Ertürk (2015) found significant variation among Pinus brutia Ten. provenances for drought resistance. In Malaysia, Ginwal and Gera (2000) reported that Acacia nilotica from Dohlpur provenance performed well in terms of growth and survival, and appeared to be useful for various types of plantation work.

Improved understanding of the diversity of T. superba,P. angolensis, P. macrocarpa and M. altissima within and between the ecological regions of C?te d’Ivoire could benef it afforestation and agroforestry programs by identifying adapted and resilient genotypes. In this context the seeds of established, mature trees become a valuable commodity (Chauhan et al. 2016). Their storage and germination requirements, and their growth characters and adaptations should be documented. These types of studies have not yet been completed in C?te d’Ivoire so we undertook this study to achieve the objectives of documenting seed germination requirements and seedling growth monitoring. Our study goal was to identify methods that could enhance afforestation success using these species. We report here our analyses of the effect of seed provenances and their pretreatment on the percentage of germination and plant growth. We first assessed seed morphometric traits, then evaluated germination rates, and finally measured seedling growth.

Materials and methods

Materials

We studied four commercial forest trees species: P. angolensis (Welw.) Warb., 1895; Terminalia superba, Engl. and Diels, 1900; M. altissima, (A. Chev.) A. Chev, 1912 and P. macrocarpa K. Schum, 1900. These species are native to the wet tropics and semi-deciduous forest, and are used mainly for timber. Three semi-deciduous provenances of C?te d’Ivoire, viz. Daoukro (Eastern), Akoupé (South) and Daloa (West center), were investigated (Fig. 1). Seventy mother plants from four localities of Daoukro (Lékikouadiokro, Kongoti, Allocokro and Anoumabo), three localities of Daloa (Gregbeu, Gonaté and Daloa) and from the protected forest of the BESSO at Akoupé were sampled.We harvested the dry fruits and removed the seeds between 1 January and 30 March 2018 (Table 1). The seeds were kept in a cooler at 25 °C in the laboratory. A total of two hundred and seventy seeds were sampled from the pooled seeds collected per provenance and per species. These were tested for germination.

Climatic data of the study site

The studies were carried out at Jean Lorougnon Guédé University, Daloa (C?te d’Ivoire) from April to November 2018. Monthly average temperature was 26 ± 0.8 °C(range 25-27.2 °C by month). Minimum daily temperature ranged from 18.9 to 20.8 °C and maximum daily temperature was 32.3-35.7 °C by month. Relative humidity averaged 79.0 ± 3.3%, ranging from 73 to 84%. Minimum and maximum relative humidity ranged from 42 to 55% and 95 to 97%, respectively. Monthly rainfall ranged from 45 to 242 mm, averaging 133 ± 74 mm. Annual rainfall during 2018 was 1302 mm. These climatic data fit well with those of the region and the country (Halle and Bruzon 2006;Anonymous 2019).

Germination test

The enhancement of seed germination in trees species is generally achieved using high or moderate temperature,cool, warm or hot water, or scarification (Tarrega et al. 1992;Sharma et al. 2008; Oliveira et al. 2013; Kim and Han 2018;De la Cuadra et al. 2019). In this study we applied four treatments: (1) T0 was the control and included untreated seeds;(2) T1 and T2 seeds were soaked for 24 and 48 h, respectively, in water at room temperature (25-26 °C) based on the works of Nadeem et al. (2017), Esmaeilpour and Van Damme (2016), Zhang et al. (2015), Soriano et al. (2011)and Mng’omba et al. (2007) which found best seed germination in trees species in these conditions; T3 and T4 seeds were immersed in 100 °C water for 12 h and 24 h, respectively (Sharma et al. 2008; Usman et al. 2010; Fredrick et al.2016). The bags containing the seeds were kept in the laboratory at room temperature of 25-26 °C. After these treatments all seeds were sown in organic soil on 24 April 2018.

Fig. 1 Localization of the seeds provenances (Daloa, Akoupé and Daoukro) in the semi deciduous region of C?te d’Ivoire

Table 1 Provenances and periods of the seeds harvested of studied species

Experimental design

The experimental design consisted of a 3 × 4 × 3 × 4 factorial plot: 3 blocks × 4 treatments × 3 provenances × 4 species.The blocks were set up in a nursery. Each block included the three provenances (R1, R2 and R3) side by side (Fig. 2).Within each block, provenance, and species, the pots filled with organic soil were grouped by fifteen corresponding to a treatment. The seeds were sown two per pot on 24 April 2018.

Fig. 2 Experimental design in nursery showing one block with the three provenances (R1, R2 and R3) line up side by side

Nursery care

Pots were watered daily and weeds were manually removed when they appeared. Insect infestations were treated by using Decis Forte?100 EC-Bayer insecticide in a mix of 8 mL Decis in 16 L of water.

Parameters assessed

The parameters assessed were morphometric characteristics and germination rates of seeds, and growth rates of seedlings. We recorded seed length, width, thickness,mass, and tegument thickness and hardness of ten healthy seeds per species. Seeds were incubated for 72 h at 70 °C in a drying oven and then immediately weighed (González-Rodríguez et al. 2011). The sizes were measured using a caliper Steel Core 6 in. Stainless Steel Electronic Digital with LCD Display.

For all seeds we recorded the duration in days from planting to germination. We calculated the time to reach maximum germination for each trial of ten provenance seeds. We considered a seed as germinated when the cotyledons emerged from the soil. Data were collected every 2 days from the first germination date until 60 days after sowing.

For seedlings we recorded height, stem collar diameter,number of foliar internodes, and leaflength and width. We measured ten plants at 6 months after germination per species, provenance, and block. Only, the T0 (control) plants were studied since this treatment yielded sufficient plant numbers 6 months after the seedlings were transplanted into pots.

Data analysis

The statistical analyses were computed using two-way ANOVA for the comparison of provenance and species effects on seed morphometric. For seed germination and seedling growth, the block effect was first tested using oneway ANOVA. In the case of no significant effect, the data were pooled. Then three-way ANOVA was used to test the effect of provenances, treatments and species on germination, while two-way ANOVA was applied to deal with the influence of the provenance and species on seedling growth.Post ANOVA Newman and Keuls test was used to compare means of provenances, treatments and tree species. The significance level was set atp= 0.05.

Results

Seed characteristics

The means differed significantly between the provenances within each species except for P. macrocarpa (Table 2). For P. angolensis and M. altissima the provenance R1 (Daloa)and R2 (Daoukro) produced the biggest seeds. While for T.superba R1 and R3 had the greatest mean seed length.

Seed hardness was moderate for P. angolensis, M.altissima and P. macrocarpa. Terminalia superba seeds were hardest and thickest (0.8-1.3 mm). Seed thickness varied more within than between provenances for all species.

Pycnanthus angolensis and P. macrocarpa seeds were largest while those of T. superba and M. altissima were smallest (Fig. 3). The seeds of T. superba were thicker and harder than those of other species.

Table 2 Studied trees species, plants provenance and averages values of the dry seeds characteristics

Fig. 3 Seeds of a Pycnanthus angolensis, b Terminalia superba, c Mansonia altissima and d Pterygota macrocarpa

Seeds germination dynamic

We recorded no germination for P. angolensis. For the others species, germination began 15-20 days after sowing (DAS). No significant treatment or species effects were recorded. The time to reach maximum germination was 30 DAS for T. superba and M. altissima (Fig. 4 a and b), and 55 DAS for P. macrocarpa (Fig. 4 c). Neither provenance nor seed characteristics affected germination dynamic of any species.

Treatments effect on germination

The percentage of germination for the three blocks at 60 DAS did not vary significantly (p = 0.91) by treatment or provenance. Therefore, data were pooled for the three blocks and then analyzed using three-way ANOVA to assess treatment, provenance, and species effects.

Treatment means differed significantly at 60 DAS for all speciesp= 0.000 (Table 3). Among T. superba trials,germination rates for T0, T1, T2, T3 and T4 were 73%,63%, 53%, 3% and 0%, respectively. T0 and T1 did not differ, nor did T1 and T2. Treatments T0, T1, and T2 were superior to treatments T3 and T4.

For M. altissima treatments T3 and T4 did not favor seed germination. Treatments T0, T1 and T2 germinated at statistically similar rates of 90%, 87% and 73%, respectively. These three treatments were superior to T3 and T4.

Pycnanthus macrocarpa treatments followed the same pattern, with T0, T1 and T2 yielding highest germination rates at nearly 90%. For T3 and T4 the rates were 3% and 23%, respectively.

T0, T1 and T2 were consistently the best treatments.They yielded 90% germination in P. macrocarpa and M.altissima versus 70% for T. superba.

Fig. 4 Seeds germination rate evolution for each treatment in a Terminalia superba, b Mansonia altissima and c Pterygota macrocarpa

Seed provenance effects on germination rate

Germination rate was affected by provenance for all species(p = 0.000; Table 3). For M. altissima the provenance R1 had the lowest germination rate for all treatments: less than 25% versus more than 50% for R2 and R3 (Fig. 5 a). For P.macrocarpa R3 yielded the lowest germination rate at less than 30% (Fig. 5 b).

For T. superba the provenance effect was significant only for T1 which yielded a low germination rate for Daoukro(R2) at 20% as compared to nearly 50% for R1 and R3(Fig. 5 c).

The often contrasting relationships between provenance and seed size indicated that the provenance effect ongermination was independent of seed size. Indeed, R1 and R2 which had similar seed characteristics for M. altissima differed greatly in germination rates. R2 and R3 for P. macrocarpa followed this same trend. In contrast, for T. superba larger seeds (R2 and R3) yielded higher germination rates.Finally, no clear relationship was established between seed characteristics and germination rates.

Table 3 Three-way-ANOVA relative to the effect of the seeds provenances, the trees species and the treatments on the seeds germination rate

Fig. 5 Averages seeds germination rate according to the provenances in a Mansonia altissima, b Pterygota macrocarpa and c Terminalia superba

Provenance effect on plant growth

Plant growth parameters were similar for all three blocks(p> 0.05). Therefore, the data were pooled and the analysis was based on 30 plants per provenance and per species.

Growth parameters differed significantly by provenance for species except P. macrocarpa. The tallest T. superba plants (55 cm) with largest leaves 22.2 cm × 8.4 cm were recorded for provenances R1 and R3 (Table 4). Nevertheless,growth varied widely within each provenance as exemplified by plant height (Fig. 6 a). For M. altissima seedlings from provenances R2 and R3 grew tallest (average 40 cm height)and grew the largest leaves (19 cm × 7.6 cm) (Table 4). Plant height varied widely in R1 and overlapped that of R2 and R3 (Fig. 6 b). The difference between provenances was significant only for the leaf size in P. macrocarpa (Table 4).The similarity in plant height distribution of the three provenances is shown in Fig. 6 c.

Terminalia superba seedlings grew faster than those of M. altissima or P. macrocarpa. The two latter species had similar growth rates.

Discussion

Our results can be summarized as three outcomes: (1) the dynamic of seeds germination was similar in the studied species whatever the treatments and the provenance, (2) seeds germination rate depended on the treatment applied and (3)the provenance impacted considerably the germination rate and the plants growth.

Time to germination (TBG) and time to reach maximum germination (TMG) were two important parameters of seed germination. These parameters were considerably influenced by seed treatments, conf irming the report of Asomaning et al. (2011) for T. superba in which TBG was 12-30 days,depending on temperature. TBG also varied between trees species as reported by Toledo-Aceves (2017) between Fraxinus uhdei (TBG = 8 days and TMG = 49 days) and Sideroxylon contrerasii (TBG = 122 days and TMG = 434 days).TBG and TMG in our study were short and their variation by species was also short. That could easier the colonization of news areas as reported by L?nnberg (2012) and Turnbull et al. (2012).

Fig. 6 Plants height distribution within the provenances R1, R2 and R3 in a Terminalia superba, b Mansonia altissima and c Pterygota macrocarpa

This study revealed that untreated seeds or those kept in immersion in water at room temperature produced higher germination rate in comparison with those treated using hot water. Hence, the high temperature and the long exposure time (12 and 24 h) probably damaged the embryo. We suggested in future work to reduce the exposure time to for example 5-15 min as done by Sharma et al. (2008) and Usman et al. (2010), since these authors found great germination at these times.

The germination rates recorded in this study were high(70-90%) compared to 30% obtained for Ceiba pentadra(Soriano et al. 2011; Nakar and Jadeja 2014). However,study could be carried out in T. superba to increase the germination rate other the 70%. Thus, organic manure, like as poultry droppings (Agera et al. 2019) as done in Eucalyptus camaldulensis could be tested.

The absence of germination in P. angolensis was much surprising since the seeds were harvested at the same period and stored in the same conditions like as the others species.Meanwhile, this result could be due a fast loss of the germination capacity. This phenomenon is well known in oleaginous seeds. More investigations should be done on P. angolensis seeds conservation, sown period and pre-treatments.

The significant effect of provenances on germination rate found in our study could be related to the trees genetic background which determined the quality and the germination of the seeds (Bagchi et al. 1990). Also ecological factors such as temperature or rainfall prevailing during the fructification could be involved in poor or high percentage of germination as suggested by Humara et al. (2000) in Eucalyptus nitens .Therefore, investigations about the temperature, rainfall and soil of the studied provenances should be achieved as done by (Kolodziejek 2017) in Peucedanum oreoselinum. Furthermore, studies like as provenances trials should be undertaken as carried out by Ginwal and Gera (2000) in Acacia nilotica.That would guide provenances selection

The percentage of germination in the different provenances in relation with the seeds morphometric characteristics seemed to indicate that larger seeds resulted in high germination in T. superba and while by contrast in P. macrocarpa and M. altissima no clear relationship was obtained. Thus the situation depended on of each species.Fornah et al. (2017) found in Gmelina arborea that large and medium seed sizes had significantly higher germination rate than the small seed size. The seeds size variations between provenances could be the result of ecological adaptation as also postulated by Fenner and Thompson (2006), Elmagboul et al. (2014), Fredrick et al. (2015) and Olejniczak et al.(2018). A genetic differentiation also might be involved(Kolodziejek 2017). In our case better understanding was required. We should study the germination of different seeds mass class (small, medium or large) as done by Fornah et al.(2017) in Gmelina arborea to comfort our results.

The assessments of the young plants growth revealed also a significant provenances effect, markedly in T. superba and M. altissima. Interestingly strong growth variability was pointed out within some provenance. Such variability within and between provenance could resulted from genetic effect as it has been suggested also for fitness-related trait by Bischoffet al. (2010). The differentiation between the provenances was a key result allowing collection of seeds or plants from provenances that showed seedlings best growth.However, provenance trial should be set up and observations upon adult plants were required to comfort our results.

Table 4 Studied trees species, plants provenance and averages morphological traits values

Conclusions

We showed in this study that seeds germination in T.superba, Pterygota macrocarpa and Mansonia altissima did not involve complex technology. They could be sown into pot directly or after soaking during 24 or 48 h in water at room temperature since these treatments promoted higher percentage of germination. Seeds germination was rapid and great. The dynamic of germination did not differ significantly between the species and was not influenced by treatment or by provenance. For P. angolensis in which no germination was recorded, study could be dedicated on seeds conservation, sown period and pre-treatments for more understanding. Another important outcome of this study concerned the differentiation between the provenances for the seeds morphometric characteristics as well as for the percentage of seeds germination as for the plants growth. Some provenances clearly exhibited larger seeds, higher germination and growth rate than the others. Thus this study lays the foundation of provenances selection. Seeds or plants from provenances exhibited best growth should be collected for production of seedlings to be transplanted onto reforestation sites. Provenance trials and molecular analysis should be carried out to get more knowledge about the provenance effect. Such studies would guide the selection and help the program of afforestation and agroforestry in C?te d’Ivoire.

Author’s contribution Sélastique Doffou Akaf-fou set up the experimental device and wrote the manuscript. Aimé Kouassi Kouame set up the experimental device and collected the data. Nestor Bi Boh Gore and Georges Yao Abessika brought help in data collection. Henri Kouadio Kouassi,Perla Hamon, Sylvie Sabatier and Jér?me Duminil corrected the manuscript.