Romeet Saha · Harish Singh Ginwal ·Girish Chandra · Santan Barthwal

Abstract Clonal propagation of eucalypts has considerable importance due to the increasing demands for short rotation tree crops. Rooting quality is important as it governs the soil exploitation capacity of the plant and the anchorage of trees which are susceptible to wind damage. This study assesses the quality of adventitious rooting by coppice cuttings of commercially important Eucalyptus clones using multiple attribute ranking of the differences in parameters of root growth. The effects of different concentrations of indole-3-butyric acid (IBA) on root development were observed.Cuttings were treated with 500, 1000, 4000, 6000 mg L -1 IBA and tannic acid for 10 s, 2 h and 24 h. Total length of root systems, number of roots, shoot to root ratios, number of root segments, extent of forking, rooting percentage, average root diameters, and number of root tips were measured.Grey relational analysis was used to create a comparability sequence to rank treatments. Reducing the concentration of auxin (IBA) and increasing the length of exposure produced better quality roots in Eucalyptus camaldulensis and interspecif ic hybrids (reciprocal hybrids of E. tereticornis and E. camaldulensis), while the opposite was observed with E.tereticornis clones. A root quality index was proposed, based on the Dickson quality index for the assessment of root system characteristics and considered total mass, shoot: root ratios, total length of root systems, and average root diameters. It has the advantage of implementation convenience.A positive correlation was obtained between grey relational analysis grades and root quality index. Rooting dynamics were studied by evaluating the total length of the root system at seven-day intervals and plotting daily current and medium increments using regression analysis. The curves showed the variation in growth rates among the different clones, and their intersection gave the optimal time of permanence (time at which further growth is restricted) which varied considerably. The highest daily current increment was 35—40 days for all clones.

Keywords IBA · Eucalyptus · Rooting dyna mics · Root quality · Grey-relational a nalysis

Introduction

With the advent of industrialization there has been considerable demands on natural ecosystems and plantations to address the demand for timber and non-timber forest products. This has given rise to domestication and cultivation of short rotation species. Eucalypts, introduced to India in 1790 from Tasmania, Australia, have been widely adopted as pioneer species in producing raw material for paper and pulp industries (Varghese et al. 2008).Eucalyptus tereticornisSm. andE. camaldulensisDehnh. candidate plus tree (CPT)clones are mainly used for plantations in India (Kulkarni 2015).Eucalyptusspp. have been selected for this study due to their industrial importance and an assessment methodology for the evaluation of rooting quality of propagated cuttings is highly relevant from an economical viewpoint. In addition,Eucalyptusspp., being canopy dominant/sub-dominant are susceptible to uprooting or trunk breakage due to high velocity winds (Williams and Douglasf 1995). Therefore the development of a quality root system should be prioritized during selection for outplanting. Clonal or vegetative propagation by cuttings is widely used for eucalypts due to low cost and relatively easy handling of materials(Titon et al. 2006). It has the added advantage of preserving the desired parental traits and replicating them in the new plantlets. However, any self-improvement that may occur in the course of sexual reproduction is absent in vegetatively propagated plants. Another reason for vegetative propagation instead of propagation by seed, was that during 1982—1995,seedling plantations were severely affected by a foliar blight disease caused byCylindrocladiumspp., and large-scale seedling mortality due to termites (Kulkarni 2013) caused severe losses. In such cases, superior clones with diverse disease and stress resistant characters are preferred.

The juvenility or freshness and vitality of the plant plays a major role in the success of rooting, thereby presenting diff iculties in obtaining clones of candidate plus trees (CPTs).Ontogenic aging or developmental changes results in maturity in plant tissues and becomes a deterrent in adventitious root induction (Corrêa and Fett-Neto 2004; Baccarin et al.2015). However, to counteract such difficulties and retain juvenility, techniques such as micro-cuttings (Assis et al.1992) and mini-cuttings (Xavier and Wendling 1998; Wendling et al. 2000) have emerged. In India, coppice cuttings are widely used as the main propagule or vegetative structure for cloning. Coppicing is a simple, inexpensive, and reliable method to obtain juvenile material for clonal propagation(Dhiman and Gandhi 2014). However, careful consideration must be given to the coppice age because coppiced shoots loose rooting ability with increasing stump size due to intershoot competition (Leakey 2014). However, coppicing is an excellent option as rooting ability has been attributed more towards physiological than ontogenic aging. Environmental conditions are optimized and rigorously controlled in propagation chambers but despite all efforts, low rooting percentage results in high economic losses to industries (Husen et al. 2017).

It is widely accepted that exogenous auxin application dramatically improves the rooting ability of cuttings, and studies have indicated that indole-3-butyric acid (IBA) has been the most successful auxin in promoting root initiation,induction and expression (Wendling et al. 2000; Almeida et al. 2007; Schwambach et al. 2008; Brondani et al. 2012;Husen et al. 2017). However, high IBA concentrations inhibit anatomical changes in adventitious root formation(da Costa et al. 2013). Post-severance physiological stress can be minimized by utilizing mist propagators, effectively decreasing water loss by providing a humid environment(Leakey 2014) and helping to acquire root effectiveness(Druege et al. 2019). Vegetative propagation under intermittent misting by the Indian Tobacco Company (ITC) Ltd,India, produced 70% cuttings with roots (Kulkarni 2015)but the rooted cuttings were still susceptible to diseases, not only to nursery conditions but the outplanting of surviving seedlings carry the possibility of disease expression in plantations. Therefore, optimizing the time cuttings are subjected to high humidity is signif icant.

Difficulty in measuring root growth have resulted in widely correlating outplanting survival and productivity with shoot morphological characters (Davis and Jacobs 2005; Thompson 1985; Bayley and Kietzka 1997; Bouma et al. 2000; Costa et al. 2001; Jacobs et al. 2005). Aboveground morphological measurements have been inadequate for the prediction of outplanting performance; therefore the need to assess root quality at an early stage is critical (Davis and Jacobs 2005; Grossnickle 2005). Root system f ibrosity, measured as the length of laterals and root tips (Deans et al. 1990), is considered a root quality parameter, similarly, f irst-order lateral roots have also gained acceptance as a useful parameter (Kormanik 1986; Dey and Parker 1997;Ponder 2000; Ward et al. 2000; Noland et al. 2001; Davis and Jacobs et al. 2005) but these attributes do not encompass the entirety of the root system. Root length in most studies is based on the line intersect method (Tennant 1975), relying on manually counting grid line-root intercepts which are laborious and prone to human error. Root system area is considered accurate but highly time- consuming (Thompson 1985), but with advances in image analysis, it has the potential to become a critical quality evaluation parameter(Rigney and Kranzler 1997). Adventitious root quality has been studied on the basis of number of roots per cutting, root length, rooting percentage and survival percentage (Sabatino et al. 2014), but the criteria were studied individually.Branching frequency/forking and the number of segments the root system is divided into determine a high frequency of higher order root proliferation, ensuring a better and more extensive exploitation of soil volume (Henderson et al. 1983)and mean root longevity (Atkinson et al. 1999), thus making forking and segmentation important parameters for root quality assessment. Average root diameter is another excellent parameter because it determines the soil penetration ability of a plant (Misra et al. 1986).

Complicated interrelationships between root develop parameters requires a study of adventitious rooting based on multiple performance attributes in an integrated manner of evaluation (Kuo et al. 2007).

Considering these aspects, this study (1) assesses the parameters governing adventitious root development in coppice cuttings of commercially importantEucalyptusclones;(2) developes an adventitious root quality index (RQI); and,(3) evaluates the optimal time for rooted cuttings kept in intermittent misting conditions.

Materials a nd methods

Choice of c lones

Six commercial clones ofEucalyptusspp. were used in this study. Clones 1, 72, 288 and 411 were purchased from“Pragati Biotechnologies”, Hoshiarpur, Punjab, while clones FRI 4 (E. tereticornis×E. camaldulensis) and FRI 5 (E. camaldulensis×E. tereticornis) were obtained from the Central Nursery, Forest Research Institute, Dehradun.These are selections ofE. tereticornis(clones 1, 72 and 288) andE. camaldulensis(clone 411) (Lal et al. 2006).All the clones in this study were selected after careful consideration of their economic importance specif ic to Indian conditions. All four commercially purchased clones are considered to be some of the most important and adaptable clones (Kulkarni 2015). Clone 1 is highly adaptable to black soil while clone 411 shows high productivity on saline sandy soils. Clones 1, 288 and 411 are resistant toEucalyptusgall caused byLeptocybe invasa.Clones 1 and 288 display clear bole and disease resistance while clone 411 is adaptable to refractory sites (Kulkarni 2015).FRI 4 and FRI 5 are superior interspecif ic hybrids ofE.tereticornisandE. camaldulensisof the Forest Research Institute, Dehradun. Extensive f ield trials have shown much higher productivity compared to their parent material (Venkatesh and Sharma 1977). Lal et al. ( 2006) specif ied clones 1, 72 and 288 asE. tereticornisand clone 411 asE. camaldulensis.Based on these characters, six clones were chosen for this study and the ramets planted in the Forest Research Institution’s Division of Genetics and Tree Improvement nursery. The soil in the plot was sampled at eight locations, bulked together for a composite sample, and submitted to the Indian Institute of Soil and Water Conservation for soil characterization. Total organic matter and total nitrogen were 0.86% and 0.10%,respectively. Available phosphorus and potassium were 10.8 ppm and 102.1 ppm, respectively. The soil had a pH of 5.3 with a f ine sand component of 49.5%, with clay,silt and coarse sand at 23.2%, 20.0%, and 7.3%, respectively. The plants were irrigated daily with 5 L water, and a bimonthly application of chloropyriphos 20% EC (emulsif iable concentrate), and a monthly application of carbendazim 50% (wettable powder) and a monthly spraying of Dithane? M45 (a contact fungicide belonging to the dithiocarbamates group). After 18 months, the outplanted ramets were coppiced to induce a f lush of juvenile shoots and to counter ontogenic aging. To minimize inter shoot competition proposed by the shoot competition hypothesis(Leakey 2014) and to maintain the juvenile gradient of the coppice shoots, only the lower most shoots were pruned

Preparation o f cuttings

Before collection of cuttings, gas-exchange parameters were measured 09:00 to 11:00 am using the CIRAS-3 photosynthesis system on fully expanded ramet leaves. A 4.5 cm2area was covered by the cuvette and photosynthesis rate(A) (μmol m-2s-1), transpiration rate (E) (mmol m-2s-1),intercellular CO2concentration (Ci) (μmol mol-1) and stomatal conductance rate (gs) (mmol m-2s-1) were measured.Gas exchange parameters were determined to assess physiological activity. Ambient conditions used for estimating gas exchange parameters and measurements were recorded after a steady state was achieved. Two types of cuttings, uniand bi-nodal, were prepared from juvenile coppice shoots(Fig. 1). Nodal variations in number of nodes retained is an experimental variable to examine rooting ability. The purpose of using uni-nodal cuttings as well as only the more common bi-nodal variant is to obtain a larger number of cuttings from the stock plant, thereby reducing overall costs.The size of the cuttings were 5.0 ± 1.0 cm) in order to maintain similar storage capacity of assimilates. The upper pair of buds were retained and leaf area was cut up to 50% to reduce transpiration. The cuttings were immediately immersed in water to prevent desiccation and all materials were prepared early morning and under shade protection to reduce mortality. The basal region of the cutting was subjected to 0.2%bavistin for 10 s before application of treatment to minimize fungal attack.

Optimization of e xogenous I BA appl ication and development of RQI

Seven treatments were administered: IBA Quick Dip(immersion of basal region in IBA (HiMedia) for 10 s)at 4000 and 6000 mg L-1; IBA Prolonged Dip (immersion of basal region in IBA solution for 24 h) at 500 and 1000 mg L-1; IBA Pulse Treatment (immersion of basal region in IBA solution for 2 h) at 4000 and 6000 mg L-1;and IBA (4000 mg L-1) + tannic acid (2000 mg L-1) for 10 s.

Fig. 1 a Uni-nodal cutting; b bi-nodal cutting

The treatments were prepared by dissolving IBA in alcohol(Absolute Ethanol, 99.9%) and the volume adjusted by adding distilled water (9:1, water: alcohol, v/v) immediately before application. After auxin treatment, the cuttings were inserted in 150 cc root trainers to a depth of 2 cm f illed with sundried, pre-soaked vermiculite in water, misted for 35 days to a relative humidity of 85% ± 2%. Maximum and minimum diurnal temperatures were 32 ± 2 °C and 28 ± 2 °C, respectively. After 35 days, the cuttings were transferred to a glass house for hardening offfor 25 days, after which measurements were taken using Biovis Image Plus (Expert Vision Labs). Measurements included number of roots, weight of rooted cuttings, shoot: root ratios, total length of root system(TLRS), number of divided segments of the root system,forks (branching frequency), rooting percentage, average root diameter, total area of the root system, and number of root tips. A modif ied “Dickson quality index” was used to formulate an adventitious root quality index (RQI; (Dickson et al. 1960). The original quality index considered the ratio of total mass (TM) to the sum of the S/R (shoot/root ratio)and the sturdiness quotient (SQ). The latter was originally the ratio of shoot length to shoot caliper diameter (Dickson et al. 1960; Currey et al. 2013). However, the purpose of this study was to assess root quality; therefore, the equation was modif ied as:

where, RQI is the root quality index for adventitious roots;TM= total mass of the rooted cutting (gm);S/Rthe shoot to root ratio andRSQthe root sturdiness quotient (SQ has been modif ied to evaluate root sturdiness and is average diameters of roots/total length of the root system).

Determination of op timal time of pe rmanence

A single treatment of 4000 mg L -1 IBA was administered for 10 secs to all cuttings and rooting was evaluated at sevenday intervals up to 49 days. Each evaluation was based on three randomly sampled cuttings from each clone, and the total length of root system (TLRS) was determined. The daily current increment (DCI) and daily medium increment(DMI) were determined according to Brondani et al. ( 2012).Using the methodology of Ferreira et al. ( 2004), the optimal time of permanence (time at which further growth is restricted) was calculated as the intercept of the daily current and medium increment curves.

where,T= time of evaluation;L(t+1)= total length of root system evaluated in timet+1;L t= total length of root system evaluated in time t,T t= days in time t, DCI = daily current increment; DMI = daily medium increment

Statistical anal ysis

The experiment was a completely randomized design with triplicates for each treatment and each clone in the case of bi-nodal cuttings, and 25 replicates in the case of uni-nodal cuttings. The data was subjected to the Shapiro—Wilk test for normality of data and the Lavene’s test for homoscedasticity of variance which were signif icant and therefore an alternative to ANOVA, i.e., Kruskal—Wallis test for k-sample independent analysis was employed. For callus formation,rooting was considered to be 0, while the missing value technique was applied in the case of mortality. The means of the parameters were subjected to pair-wise analysis at a 5% level of signif icance as a post hoc test. SPSS software(SPSS for Windows v23, Chicago, Illinois, USA) was used to process the data.

Thereafter, grey relational analysis (GRA), proposed by Deng ( 1982), was applied to solve the complicated interrelationships between multiple characters (9 in the present case;Table 1), and multiple treatments (7 in the present case). The steps of GRA (Kuo et al. 2007) are:

Step 1 Grey relational generating (GRG)

This transforms all observed values (y ij)for each treatment into a comparability sequence (x ij)by Eqs. (1) and (2)for the-larger-the-better and smaller-the-better characters,respectively.

Table 1 Total parameters considered for GRA vs parameters considered for RQI

where,yijis the measured value ofjthcharacter of theithtreatment

No spaceYican be transformed into the comparability sequence asXi=(xi1,xi2,xi3,xi4) using Eq. (4)

In this study, the only smaller-the-better character is the shoot—root ratio while all others are the larger-the-better characters.

Step 2 Def ining reference sequence

Step 3 Determination of gray relational coefficient

To see the closeness ofx ijwithx0j, the grey relational coefficient,δ(xij,x0j) between comparability sequences and the reference sequence is determined as:

where, Δij=i=1, 2, … ,8;j=1, 2,… , 9 ,d∈ [0, 1] is an distinguishing coefficient. The purpose ofdis to expand or compress the range of the grey relational coefficient. In this study, the results are based upond= 0.5.

Step 4 Calculation of grey relational grades

Grey relational gradeΓ(X0,Xi) , betweenX0andXi,i=1,2,…,9 is required to calculate the performance of the treatments based upon the nine characters as follows:

where,w jdenotes the weight assigned to thejt hcharacter with

For the purpose of this study, certain parameters (number of roots, TLRS, shoot: root, rooting percentage) were given a weight of 2/13, while root area, segments, forks,number of root tips and average diameter of roots were kept at 1/13, maintaining a ratio of 2:2:2:2:1:1:1:1:1. This grade def ines the similarity between the reference and comparability sequence, i.e., a higher grade denotes a higher degree of similarity, thereby providing a better rank for the treatment in question.

The highest value ofΓ(X0,Xi) means that theit htreatments will be the best choice.

Results

Assessment of parameters governing root development

The gas exchange parameters showed no correlation with root growth parameters. Photosynthesis rates ranged from 7.84 to 20.65 μmol m -2 s -1 (Table 2) and the stock plants were considered to be physiologically active during the collection of cuttings due to positive rates of assimilation.

Treatments on the six clones showed highly signif icant differences (p＜ 0.01) in most cases, and signif icant differences (p＜ 0.05) in all cases of uni- and bi-nodal cuttings(Table 3). The number of roots was highest when clone 411 was subjected to an IBA Pulse treatment of 4000 mg L-1(18.3 ± 3.5), while average number of roots from 2.5—4.6 was observed in all bi-nodal cuttings (Table 4). The number of roots of uni-nodal cuttings varied from 0.3 to 3.6.With bi-nodal cuttings, IBA Quick Dip at 6000 mg L-1with clone 1 resulted in the highest number of roots (7.3).Pair-wise comparison of Kruskal—Wallis analysis showedthat this treatment produced results signif icantly different from IBA Prolonged dip at 500 mg L -1 (no. of roots = 0,p= 0.007), IBA Prolonged dip at 1000 mg L-1(no. of roots = 0.7,p= 0.031), and IBA Pulse treatment at 6000 mg L -1 (p= 0.007) (Table 3), whereas rooting was absent in the case of uni-nodal clone 1 cuttings.

Table 2 Gas exchange parameters measured immediately before collection of cuttings

I 5odal FR Bi-N 22.070 0.002 21.860 0.003 20.562 0.004.070 22 0.002 21 0.003.820.355 18 0.010 17 0.014.592.843 20 0.004 7.000 Uni-Nodal 127.861 0.000 133.481 0 0.00 119.712 0 0.00 127.058 0.000 0.000 130.404 0 0.00 132.709 131.254 00 0.00 132.024 0.00 FRI 4 7.000 al od Bi-N 21.402 0.003 19.204 0.008 19.277 0.007 02 21.4 0.003 21.457 0.003 19.139 0.008 21.714 0.003 22.377 0.002 0 7.00 i-Nodal 2 2.120 5.17 82 1 1.931 2.52 4 Un 15 0.000 16 0.000 16 0.000 15 0.000 2.84 14 0.000 8.24 15 0.000 15 0.050.0878 6.06 15 0.000 7.000 dal cuttings 11 Clone 4 odal 22.052 2 Bi-N 0.00 0.01 18.410 0 09 0 16.6 0.02 0.00 21.9 0.00 22.052 2 98 3 15 3 21.2 0.00 21.660 3 0.00 0.00 i-no 21.633 3 7.000 ni- and b in u Uni-Nodal 125.151 0.000 112.483 0.000 99.269 0.000 52 89.2 0.000 107.747 0.000 111.147 0.000 113.586 0.000 113.027 0.000 0 7.00 eters of adventitious rooting Clone 288 odal B 20i-N.294 0.005.653 20 0.004.728 16 0.019 20.294 0.005.350 20 0.005 20 0.004.912.331 20 0.005.864 19 0.006 7.000 171.750 dal Uni-No 0 0.00 171.750 0.000 171.750 0 00000 rowth param 0.00 0.00 0.00 171.750 171.750 0.00 0.00 171.750 171.750 0.00 f diff erent treatments on g 171.750 7.000 e 72 Clon al od Bi-N 42 22.5 0.002 92 18.8 0.009 17.193 0.016 42 22.5 0.002 22.759 0.002 21.169 0.004 20 21.2 0.003 75 21.4 0.003 0 7.00 Clone 01 odal B 21i-N.627 0.003.776 17 0.013.377 21 0.003 21.627 0.003.198 21 0.003.090 20 0.005.459 16 0.021.184 21 0.004 7.000 Kruskal-Wallis statistics o Chi Square ig.Asymp. S quare ig.Chi S Asymp. S quare ig.C Ahsyi mSp. S quare ig.Chi S Asymp. S quare ig.Chi Sp. Sare ig.Asym C Ahsyi mSqup. S Chi Square ig.Asymp. S Chi Square ig.p. S Asym df Parameters Test statistics ength ootsiameter Table 3 oot tss Root L No of R Average D Shoot: R Area Forks en Segm Root tip

For clone 72, IBA Quick Dip at 6000 mg L -1 (no. of roots = 3.3) and IBA Quick Dip at 4000 mg L-1(no. of roots = 4.7) did not result in signif icant differences, however IBA Prolonged dip at 1000 mg L -1 (no. of roots = 0)was signif icantly different from IBA Quick Dip at 6000 mg L -1 . With clone 288, once again IBA Quick Dip at 6000 mg L -1 (no. of roots = 5) resulted in signif icant deviation from both IBA Prolonged dip treatments (no. of roots = 0.7,p= 0.029) and IBA Pulse treatments (p= 0.002) (Table 3).All threeE. tereticornisclones (1, 72 and 288) produced greater number of roots with IBA Quick Dip than with the other treatments (Table 4). For clone 411 (E. camaldulensis), greatest number of roots was produced with IBA Pulse treatment at 4000 mg L -1 but there was no signif icant difference (p= 0.541) between this treatment (Table 3) and IBA Prolonged dip at 1000 mg L -1 (no. of roots = 11) (Table 4).However, there was a signif icant difference between IBA Prolonged dip at 1000 mg L -1 and the IBA Pulse treatment at 6000 mg L -1 (p= 0.007) and the controls (p= 0.007). I In all exogenous IBA applications with bi-nodal cuttings,with the exception of IBA Pulse at 6000 mg L-1(where callusing was the predominant feature due to IBA Prolonged exposure to high concentration of IBA), rooting percentage was higher than controls (38.9%) and was highest with IBA Quick Dip at 4000 mg L -1 (94.4%), followed by IBA Quick Dip at 6000 mg L -1 (83.3%). With uni-nodal cuttings,mortality was signif icantly higher and a range of rooting percentage of 3.0%—35.0% was achieved. Clone 1 and FRI-5 did not produce roots. However, IBA Quick Dip at 6000 mg L -1 resulted in an average rooting percentage of 58% by uninodal cuttings in the other four clones. Based on observations of the number of roots produced, bi-nodal cuttings have a clear advantage over uni-nodal cuttings of all clones tested.

Root areas were signif icantly different (p＜ 0.01) (Table 3)between uni- and bi-nodal cuttings of all clones, and ranged from 2.5—3.6 cm 2 and 2.4—8.5 cm 2 , respectively (Table 4).For bi-nodal cuttings of clone 1, the largest root area was produced by cuttings treated with IBA Pulse at 4000 mg L -1(15.4 cm 2 ), but there was no signif icant difference with IBA Quick Dip at 6000 mg L -1 (area = 10.93 cm 2 ,p= 0.597),according to Kruskal—Wallis pair-wise analysis. For clone 72, the largest root area was with the IBA Pulse treatment of 6000 mg L -1 but no signif icant difference was observed between this treatment and the IBA Quick Dip at 6000 mg L -1 (p= 0.603). Significant differences were observed between the IBA Quick Dip treatment at 6000 mg L -1 and both the IBA Prolonged dip treatments (p= 0.037 and 0.002), and with the tannic acid treatment (p= 0.009). With clone 288, the largest root area produced was with the IBA Quick Dip solution at 6000 mg L -1 . A similar trend was observed with the number of roots, wherein all threeE. tereticornisclones had better root area development when treated with IBA Quick Dip at 6000 mg L -1 . Root area in clone 411 (E. camaldulensis) was highest in IBA Prolonged dip at 1000 mg L -1 (15.7 cm 2 ) (Table 4). and was signif icantly different with IBA Pulse at 6000 mg L -1 (p= 0.001), the tannic acid treatment (p= 0.012) and the controls (p= 0.001).With clone 411, FRI 4 and FRI 5 a similar trend was again observed where IBA Prolonged exposure at 1000 mg L -1 IBA was the most efficient treatment. The larger root areas of bi-nodal cuttings showed comparatively more vigorous growth than uni-nodal cuttings.

Average root diameters also showed considerable variation between the different clones and treatments (p＜ 0.05),i.e., 0.8 mm—1.3 mm and 1.4 mm—2.8 mm for uni- and binodal cuttings, respectively (Table 4). FRI 4 deviated from the observed trend and had the highest average diameter when treated with IBA Quick Dip at 4000 mg L-1which was signif icantly different from the results of IBA Prolonged dip at 1000 mg L -1 (p= 0.023).

The total length of root systems (TLRS) was considerably higher with bi-nodal than with uni-nodal cuttings (Table 4).The average TLRS ranged from 29.1 to 90.5 cm for the different clones and the highest TLRS of 165.7 cm was produced by FRI 5 (Table 4). The highest TLRS of 122.5 cm was produced by bi-nodal cuttings of clone 1 was treated with IBA Pulse at 4000 mg L -1 (Table 4) but it was not signif icantly higher (112.9 cm) from the treatment with IBA Quick Dip at 6000 mg L -1 treatment according to Kruskal—Wallis pair-wise analysis. TLRS also followed a similar pattern as the other parameters, showing a species specif ic response to treatment.

Forking of the root systems, which relates to exploitation capability was also evaluated and variations by uni-nodal cuttings (12.2 to 111.0 forks per root system) were much less than by bi-nodal cuttings (91.8 to 264.5 forks per RS). The number of forks and segments was similar for all the clones(Table 4). With bi-nodal cuttings of clone 1, both forks(424.0) and segments (590.0) were highest with IBA Quick Dip at 6000 mg L -1 treatment. This was signif icantly different from both IBA Prolonged dip treatments and IBA Pulse at 6000 mg L -1 (p= 0.004 in the case of forks andp= 0.006 in the case of segments)]The number of root tips was highest (366.5) with the IBA Pulse treatment of 4000 mg L-1(Table 4). Both the FRI 4 and FRI 5 clones had the highest forking (430.3 and 346.7, respectively), greatest segmentation (706.7 and 495.7, respectively) and highest number of root tips (352 and 302, respectively) when treated with IBA Prolonged dip at 1000 mg L -1 (Table 4).

Table 4 Effect of different treatments on adventitious rooting by bi-nodal cuttings of Eucalyptus clones

Table 4 (continued)

Table 4 (continued)

The rooted cuttings of the six clones were further evaluated based on nine parameters to validate the results obtained, and to understand the effect of each treatment when all parameters are considered together rather than separately. Grey relational analysis was employed after scaling the results from 0 to 1. The clones exhibited a high degree of variation in response to treatments; clones 1, 72 and 288, (i.e.,E. tereticornis), had better root development when subjected to high concentrations of IBA for short periods of time (Table 4). Clone 72 cuttings treated with Quick Dip at 6000 mg L -1 and IBA Pulse at 6000 mg L -1 were not signif icantly different (p＞ 0.05), and the two treatments are considered to have produced similar results with bi-nodal cuttings. Uni-nodal cuttings of clone 288 with the addition of tannic acid showed signif icant improvement in rooting parameters (p＜ 0.05). Clone 411 (E. camaldulensis), FRI 4 and FRI5 (interspecif ic hybrids), on the other hand, showed better results when subjected to long duration(24 h) and low concentrations of IBA. In most studies, the most important parameters in determining root quality are rooting percentage, the total length of root system (TLRS),length of the longest root, and the number of roots. However,induction of a large number of roots does not ensure that all roots will undergo equal rates of cell division and achieve the desired length and branching. Similarly, a higher TLRS value does not ensure desirable root diameters throughout the length; moreover, TLRS could be achieved by a single root having high branching. Therefore, to assess root quality, several parameters must be considered. To address this issue, this study concentrated on assessing the quality of adventitious roots by considering parameters such as forking, segmentation, root area, average root and number of root tips, along with the more commonly used parameters.Therefore, combining these parameters a grey relational grade was developed which provided ranks to each treatment (Table 5 ).

Determination o f root q uality i ndex(RQI)

The RQI showed the effect of different treatments on adventitious root quality and clonal variation similar to grey relational analysis (GRA). There was a positive correlation(Spearman’s ? = 0.658) between RQI and GRA grades,validating the use of RQI as an easier alternative to GRA.However, there were differences between the two indices when comparing treatment ranks between clones. Based on GRA, the IBA Quick Dip of 6000 mg L -1 was superior but the RQI gave a higher rank to IBA Prolonged dip of 1000 mg L -1 . In four of the six clones studied, the preferable treatments had similar ranking (Tables 5 and 6). The use of RQI may still be preferable to using a robust assessment due to ease of analysis and subsequent selection.

Determination of op timal t ime of permanence

Clonal variation was prominent with regards to rhizogenesis,necessitating adjusted equations for different clones. Regardless of clone, root emergence began 14 days after the start of the experiment. The optimal daily current increment (DCI)was at 35, 40, 37 and 39 days for clones 1, 72, 288 and 411,respectively. TLRS also varied between clones. At 35 days,it ranged from 12.5 cm for clone 411—22.9 cm for clone 1. A polynomial equation of the third order was used to derive the TLRS on successive days (Fig. 2 a—d). The highest increment/day was 1.43 cm for clone 1 while the lowest was 1.22 cm for clone 411. Although there was no signif icant difference in the highest daily increment amongst the clones,the time at which it was achieved is important in calculating the optimal time of permanence. DCI and DMI (daily medium increment) intersection occurred between 51 and 57 days (Fig. 3 a—d). The optimal time was between 35 and 40 days. The shaded areas in Fig.3 a—d represent the time between the highest DCI and the intersection of the curves and can be considered as an acceptable range for removal of the cuttings from the intermittent misting environment.

i-Nodal I 5odal Un 4 (7) 0 3 (8) 0 FR 0.582 (4) 0 0.942 (1) 0 0.701 (2) 0 0.585 (3) 0 0.33 0.33 0.336 (6) 0 0.382 (5) 0 ment evelopI 4odal Uni-Nodal Bi-N eters of root d FR dal Bi-N 0.657 (3) 0.851 (1)0.879 (1) 0.679 (2)0.498 (6) 0.604 (3)0.333 (7) 0.533 (4)0.546 (4) 0.333 (6)0.673 (2) 0.333 (6)0.333 (7) 0.372 (5)0.513 (5) 0.333 (6)0.814 (1) 0.80 al Uni-No 1 (2)(3)0.807 (2) 0.59 3 (6)7 (4)0.584 (6) 0.57 3 (6)4 (1)9 (5)3 (6)iff erent param Clone 411 od 0.631 (4) 0.94 0.702 (3) 0.33 0.333 (7) 0.33 0.602 (5) 0.53 0.333 (7) 0.33 on d ased s) b bi-nodal cutting e 288 Clon odal Uni-Nodal Bi-N 3 (6) 0.426 (7)0.46 i-Nodal Bi-N 5 (5) 0.568 (6)0.467 (2) 0.573 (5)0.70 0.962 (1) 0.69 (3)6 (7) 0.589 (4)0.436 (7) 0.748 (2)0.433 (3) 0.801 (1)0.667 (4) 0.333 (8)0.48 i and thesis) of individual clones (un 0.282 (2)2 Clone 7 odal Un dal Bi-N 0.358 (6) 0.282 (2)0.282 (8) 0.282 (2)0.48 (4)0.547 (3) 0.846 (1)0.554 (2) 0.282 (2)0.715 (1) 0.282 (2)0.412 (5) 0.282 (2)0.336 (7) 0.282 (2)aren Grey relational analysis (ranks in p Clone 01 odal Uni-No Bi-N 0.451 (6) 0 0.595 (4) 0 0.831 (1) 0 0.801 (2) 0 0.451 (6) 0 0.683 (3) 0 0.429 (8) 0 g L -1 g L -1 0.466 (5) 0 g L -1 g L -1 g L -1 ip 500 m 000 m g L -1 g L -1 ip 1 ip 4000 m 000 m ip 6 cid + IB o.uick D Table 5 A Pent N Treatm rolonged D rolonged D000 m uick D000 mA 4000 m IBIBIBIBIBIBA P A Q A A QPu lse 4 ulse 6 A P T Caonnntrioc Al

Discussion

In spite of the advances in mini-cutting and micro-cutting techniques, the improvement of coppice cuttings has received little attention. However, in many developing countries where they are major sources of propagules, research in this area is of considerable economical signif i cance. Vegetative propagation captures both non-additive as well as additive genes, replicating existing traits without affecting the quality of the newly formed plantlet. It is an excellent methodology for quality control (Leakey 2014). The selection of cuttings for propagation is an important step as they are dependent on stored assimilates for survival and rooting. The juvenility or freshness and vitality of cuttings is important to improve rooting capacity (Druege et al.2019),and selection of material must be made accordingly. The stock plants from which cuttings are derived contribute, to a large extent, towards rooting success and therefore careful maintenance of stock plants is important. The shoot competition hypothesis indicates that coppice shoots of the upper part of the stump have better rooting ability than those at the bottom (Leakey 2014). Therefore, in this study, bottom shoots were pruned to provide better assimilate allocation.Post-severance factors also play an important role in rooting and it is therefore important to minimize physiological shock by decreasing time after severance and applying moisture to the cuttings. To minimize carbohydrate loss by respiring tissues and water loss via transpiration. Cuttings in this study were placed into moist vermiculite and misted.The importance of auxin in promoting adventitious roots has been well documented and used extensively for mass propagation (Schwambach et al.2008; Hartmann et al.2011;Hunt et al.2011). Auxin stimulates cell differentiation, promotes starch hydrolysis and draws nutrients to the base of cuttings (Atangana et al.2011; Leakey 2014). However, not all hardwood cuttings respond uniformly to auxin and in certain cases, the method of application has been the determining factor in adventitious root formation (Serr 1964). Auxin is, however, not the only determinant of rooting; ontogenic aging also plays a major role in controlling root formation,i.e., higher mortality during root initiation phase (Stevens and Pijut 2017). Therefore, the selection of plant material is an important step to ensure high rooting of cuttings. Thetreatment of cuttings with auxin is also an important determinant in dedifferentiation of cells and tissues and subsequent redifferentiation (Luckman and Menary 2002). Husen( 2008) determined that IBA increased the accumulation of auxin in the basal part of cuttings and helped root induction, but the concentration of auxin needed to be precise to obtain the desired rooting rate and quality (Corrêa and Fett-Neto 2004; Brondani et al. 2012). Auxin modif ies the molecular mechanism of DNA, RNA and protein synthesis in the root zone and may lead to the promotion or inhibition of root induction (Husen and Pal 2007; Komatsu et al.2011). The preference of IBA over other auxins may be due to its higher stability (Bellini et al. 2014). This study tested different concentrations of IBA and their application times on cuttings of different clones to assess the effect on root development.

Table 6 Root quality index(ranks in parenthesis) of bi-nodal cuttings of clones

Fig. 2 TLRS of Eucalyptus clones in relation to time

Fig. 3 Rooting rates of Eucalyptus coppice cuttings; daily current increment (DCI) and daily medium increment (DMI) over time

It is important to assess the quality of adventitious roots and not opt only for treatments that result in higher rooting percentage because the chances of mortality may increase during the hardening stage (Bhardwaj and Mishra 2005).A well-developed root system can help prevent transplant shock after outplanting to the f ield from relatively milder nursery conditions (Rietveld 1989). This is especially relevant with regards to rooted cuttings where the adventitious root systems are protected in a conf ined space during rhizogenesis. This protects the entirety of the secondary and tertiary roots from desiccation during hardening and transplanting (Miller 1999). Therefore, quality assessment of the root system is critical for the selection of a seedling with a superior root system. Orhan et al. ( 2015) studied the quality of adventitious rooting based on root length, weight and number of lateral roots, while Yasar et al. ( 2010) rated root quality based on number of main roots, high root length and rooting percentage. A similar study onTeucrium fruticansL.was reported by Sabatino et al. ( 2014), to determine adventitious root quality based on number of roots per cutting, root lengths, rooting percentage and survival percentage. Average root diameter is also an important parameter for assessing root quality because it def ines the ability of the roots to exert sufficient pressure to penetrate the soil strata (Misra et al. 1986), and the presence of relatively lower diameter roots provides for horizontal and radial spread of roots. Such conditions could lead to higher competition in sub- surface soil layers, whereas high penetrability of thicker roots could improve better uptake of localized nutrients and moisture(Misra et al. 1986). In the case of homorhizic or f ibrous root systems as those in the present study, secondary and tertiary roots are also crucial to the uptake of nutrients due to increased surface area (Bellini et al. 2014). In this study, the forking of roots was regarded as an indicator of higher order root formation and was considered as a “l(fā)arger the better”attribute (Kuo et al. 2007). Forking (branching frequency)and segmentation, i.e., the number of segments the root system has divided into, are critical for improving a plant’s soil exploration capabilities, thereby enhancing access to nutrients (Henderson et al. 1983) and increasing root longevity (Atkinson et al. 1999). Another important parameter considered for this study was shoot: ratio (S/R) which is an indication of the preferential allocation of carbon to either shoot or root (Madhu and Hatfeld 2013). A higher partitioning of assimilated photosynthates towards root formation in early stages of development is benef icial for adventitious root formation (Veen et al. 1991). A decrease in the S/R ratio would allow for greater exploring capability of the plant after outplanting. It is often difficult to assess S/R ratios due to difficulties in assessing sub-surface biomass (Madhu and Hatfeld 2013). However, this study provided an excellent opportunity to integrate S/R ratios in the assessment of root quality, and lower ratios indicated a higher partitioning of resources towards root system development. The IBA Quick Dip treatment resulted in a lower S/R ratio for cuttings of clone 411. With FRI 4 cuttings, the S/R ratios deviated from the expected pattern, i.e., the lowest ratios were from treatment with IBA Pulse at 4000 mg L -1 . FRI 5, however, had the lowest S/R ratio when treated with IBA Prolonged dip at 500 mg L -1 , thereby following the observed trend.

Research so far has considered different parameters as stand-alone variables for assessing root development, but the importance of multiple attribute selection has been demonstrated in the present study. The interdependence of root system properties as measured by different parameters may be a good surrogate for the selection of key characteristics (Atkinson et al. 1999). In this study, the assessment of individual parameters showed thatE. tereticornisclones responded best to IBA Quick Dip treatment at 6000 mg L-1(Table 4). However, in certain cases, there were better results of different individual parameters in other treatments as well,for example, the highest total length of root system (TLRS)was with bi-nodal clone 1 cuttings treated with IBA Pulse at 4000 mg L -1 . As seen from the results of each parameter under study (Table 1), it is difficult to select the optimum treatment based on one or more key parameters taken separately. Even though a trend was observed with cuttings of clones 1, 72 and 288 (E. tereticornis) showing good rooting ability with IBA Quick Dip, while clones 411 (E. camaldulensis), FRI 4 and FRI 5 (reciprocal hybrids) responded best to the IBA Prolonged dip treatment. Deviations of certain parameters from this trend makes the selection of treatment difficult.

Where the selection of the optimum treatment becomes difficult when parameters are taken as stand-alone measures of root development, a multiple attribute decision- making protocol will make the process easier by taking into account all the different performance attributes and combining them into a single grade value for each treatment. For example, a grade of 0.879 (rank 1) was allotted to IBA Prolonged dip at 1000 mg L -1 IBA for FRI 4 cuttings even though several parameters were optimum in other treatments. Application of grey relational analysis (GRA) considered all performance factors and resulted in a combined output for the assessment of root quality based on different treatments.From the GRA scores (Table 5) it is clear that there were species- wise and even genotypic variations in response to the different treatments.E. tereticornis(clones 1, 72 and 288) responded better to brief exposure to high concentrations of IBA (Table 4), while cuttings of FRI 4 and FRI 5 (interspecif ic reciprocal hybrids) responded well to long exposure to low concentrations of IBA (Table 4). Clone 411,i.e.,E. camaldulensis, showed good rooting under both treatments but ranking showed that lengthy exposure to low concentrations of IBA were better. Although there was variable response by uni- and bi-nodal cuttings, the general trend remained the same. Uni-nodal cuttings, however, failed to produce roots in clone 1 and FRI 5, and the majority of the parameters studied showed inferior results to those of binodal cuttings (Table 4).

Accurate assessment of rooting competence [in a genotype-specif ic manner] may be more useful economically when large-scale propagation work is required. However,where the determination of numerous parameters may not be viable or when a quantitative measurement of root quality is the objective, then an alternative method of assessment is the formulation of a root quality index (RQI). The original parameters used to determine RQI had were positively correlated with outplanting success (Pramuk and Runkle 2005; Currey et al. 2013). In this study, GRA grades and RQI values were positively correlated and therefore using RQI as an alternative to a multiple attribute decision- making system may be applicable in certain cases. But grey relational analysis (GRA) may still provide a more robust testing alternative because it also considers more parameters such as forking and segmentation which are important in mass propagation. Another benef it of GRA is its ability to integrate assessment of parameters regardless of their measurement units. RQI will be useful in cases where there are infrastructural constraints in taking measurements or if rapid assessment in required. For nursery managers without access to equipment such as an image analyzer, RQI could become an important tool for the assessment of root quality. From Table 1, it is clear that RQI is a more simplif ied version of quality assessment. With GRA, the number of parameters considered was nine as opposed to four with RQI, making RQI an easier alternative when robust testing of rooting ability is not required. However, GRA still provides a better understanding of integrated selection and is a more robust testing methodology. RQI has the advantage of being easier to formulate by using relatively easy to evaluate parameters.

Another problem affecting cuttings during root development in misting chambers is the prevalence of diseases due to the conducive environment of high humidity and temperature (Ferreira et al. 2004; Goulart and Xavier 2008;Zhu et al. 2010; Brondani et al. 2012), along with auxin degradation at the cut end (Steffens and Rasmussen 2016).Therefore, it is essential to optimize the time that propagules are kept in such conditions in order to reduce mortality.Brondani et al. ( 2012) used mathematical modeling to calculate the permanence time in Camden white gum (Eucalyptus benthamiiMaiden & Cambage) mini-cuttings and found that daily current increment (DCI) was highest after 35 days for all the clones. In this study, variations in daily current increment was among all the clones, with the highest increment between 35—40 days, while the intersection of daily current increment (DCI) and daily mean increment(DMI) curves occurred between 51—57 days. Therefore, it becomes increasingly relevant to model the [optimal time of permanence] for each genotype under mass propagation to achieve maximum output and minimize mortality. Although ideally the optimum time for the removal of cuttings from the misting chamber would be at the intersection of the DCI and DMI curves, the time between optimal DCI and the intersection would also be conducive to good adventitious root development. In this study, 35—40 days would be the optimum time to transfer cuttings from the misting chamber to a hardening chamber.

In conclusion, clonal variation was highly responsive to different applications of IBA, and optimizing the treatment methodology for specif ic clones would be highly benef icial to mass propagation. The choice of methodology should consider economic viability and implementation convenience. Variability of the results with different parameters provided multiple options for the selection of the optimum treatment. Therefore, root quality assessment, in conjunction with multiple performance factors and subsequent translation into a comparability sequence, was determined (Kuo et al. 2007). The f indings of this study indicates that coppice cuttings ofE. tereticornisrespond well to a 10 s exposure to 6000 mg L -1 IBA, whileE. camaldulensisand interspecif ic hybrids demonstrated better rooting ability when subjected to 24 h exposure to 1000 mg L -1 IBA. The proposed assessment criteria provide for the selection of better quality rooting via vegetative propagation, ensuring better survival. It is recommended using Grey relational analysis to formulate a comparability sequence as the preferred for selection. Root quality index is recommended where limitations outweigh benef its. The optimum time of moving rooted cuttings for hardening-offalso varied according to genotypes. Modeling must be genotype- specif ic to maximize production along with the development of genotype-specif ic protocols for commercially important clones. This is critical for clones that are susceptible to nursery diseases. Studying the dynamics of adventitious rooting provides an understanding for the decrease in daily increment due to conf inement (Funk 1971; Aphalo and Rikala 2003; Davis and Jacobs 2005),and unavailability of nutrients for further development. It determined that 35—49 days is the optimal time for removing rooted cuttings from mist propagators. Therefore, it is recommended that the optimum time of permanence be determined separately for all commercially important clones that are widely propagated in order to minimize losses at the nursery stage. These measures are critical to the provision of higher economic returns and to ensure viable planting material for propagation.