Livestock Research for Rural Development 27 (1) 2015 Guide for preparation of papers LRRD Newsletter

Citation of this paper

Modeling the regrowth of the native Salsola vermiculata L. range species by grazing simulation

Yousfi Mohamme and Azzouzi Blel1

Institut des Sciences Vétérinaires et Agronomiques, Université Hadj Lakhder, Batna, Algérie.
yousfimohammed667@yahoo.fr
1 Faculté des Sciences de la Vie et de la Nature, Université Ziane Achour, Djelfa, Algérie.
Institut des Sciences Vétérinaires et Agronomiques – Université Hadj Lakhder,Batna.

Abstract

Foreseeing the level of production and predicting its response for future exploitation are essential elements in the management of pastoral resources over the semi arid zones. The search of the tendency of production along with the establishment of mathematical models governing the behaviour towards grazing, are essential to explore alternatives for a rational exploitation of natural resources of steppe Rangelands. The main aim (purpose) of the present study is to investigate mathematical models which can be able to determine the level of the regrowth of phytomass and dry matter of a fenced rangeland of Salsola Vermiculata L. As treatments, three levels of defoliation (severe, moderate and light), were applied to plants during three growth phenological phases ( early, full, end) with respect to the measurement of the height, the biggest and the smallest width of plants. This study showed that the rates of regrowth are highly influenced by defoliation level and the phenological stage. The rates of regrowth of the phytomass and the dry matter are following mathematical models depending on the severity defoliation degree. The phytomass rate regrowth can well be expressed by neperian logarithm equations for heavy, light and mixed defoliation treatments while the moderate treatment is expressed by a quadratic equation. For the heavy and light treatments, the dry matter regrowth rate can be expressed with logarithmic equations whereas the quadratic form of equations is appropriate for moderate treatment. For mixed treatments the hyperbolic equations can be used.

Key words: defoliation, mathematical model, phenological stage, rangeland, regrowth rate


Introduction

The last decades have seen a wide debate on the descriptive aspects of the degradation of steppe rangelands and the threat of desertification resulting from this process (Nedjraoui 2004, Nefzaoui & El Mourid 2008). But look on the power and mechanisms of regrowth of these rangelands are less discussed and, consequently, less known. In this context, emerges a new way based on the use of mathematical modeling to better understand the dynamics grazing - regrowth of pastoral resources. In addition to its benefits of saving time and money (Ferchichi 2004, Bartolome 2006), this approach in the exploration of grazed ecosystems could produce predictive tools and support decision-making, if the studies are well conducted. (Ferchichi 2004, Heitschmidt & Walker 1996, Richardson et al. 2007, Squires 1998, Wood 2004). Our study is a small step in the attempt to implement a scientific approach to understand the response of Algerian natural steppe rangelands to the effect of different degrees of severity of grazing in different phenological stages and their impact in terms of regrowth and recovery of their productive potential as implemented by many authors (Al Hartani & Fogel 1998, Bartolome 2006, Batabyal & Godfrey 2002, Salah Tag El Din 1994, Westboy et al. 1989).

The study, therefore, focused on the simulation of grazing by three defoliation degrees corresponding to three levels of severity:

Severe defoliation where 75% of the volume of the plant are harvested.

Moderate defoliation on 50% of the volume of the plant.

Light defoliation covering 25% of the volume of the plant.

These three degrees of defoliation are performed simultaneously during three phenological stages: early growth stage (April 2012), full growth stage (July 2012) and the end of growth stage (November 2012) of Salsola vermiculata L. forming the dominant species of the Darmoune rangeland situated in Thlidjène Commune (Wilaya of Tébessa) in the steppe zone of eastern Algeria.

Measurements of height (Hi), large width (LWi) and small width (SWi) were performed for each type of cut for each phenological stage and were used to quantify the volumes of each cut and dry matter corresponding thereto.

During the following season and at the corresponding phenological stage, new measures after regrowth were performed: height (Hre), large width (LWre) and small width (SWre). They allowed deducting an assessment of regenerated volumes of cut plants and the amount of dry matter thereto. The regrowth rate of phytomass (RPre) and dry matter (DMre) are subsequently established to serve in the search for appropriate mathematical models related to each treatment.


Material and methods

Study area

The study was conducted in the fenced Darmoune rangeland located in Thlidjène Commune (south of Wilaya of Tébessa in the steppe zone of eastern Algeria). The rangeland, which covers an area of 900 ha, is considered as the most famous region of sheep livestock breeding.

Climatically, the area belongs to the semi-arid climate whose interannual average rainfall is around 250 mm, while the minimum temperatures reach -4°C and maximum 40°C. While soil conditions are dominated by a skeletal loamy soil with a depth not exceeding 35 cm on average.

For long time the overgrazing coupled to practice of illicit tillage and harshness of climate conditions caused a severe degradation of the rangeland natural resources increasing the desertification risk in the region. Nowadays, the use of the rangeland is regulated under control of the High Commissioner to Development of the Steppe, a state institution attached to the Algerian Ministry of Agriculture.

Plant material

The plant material used in this study is the Salsola vermiculata L. growing spontaneously in Darmoune rangeland and forming the dominant perennial species in this rangeland. The choice of this species is a form of contribution to the understanding of the behavior of indigenous species steppe rangelands that have been none or little studied despite their importance on pastoralism in Algeria.

Experimental design

The experimental design was set up along an east-west transect of the Darmoune rangeland. It consists of random blocks, each one consisting of 03 degree defoliation levels and three levels of phenological stage and consequently 9 units per block. Treatments are repeated 05 times.

Each unit is 04 m2 of area and the units are separated by a distance of 01 m. This area is considered sufficient to contain a maximum number of Salsola vermiculata L. plants to allow experimentation and get closer to the reality of the rangeland.

Observations and analysis
Samples dimensions measurement

The choice of using the volume of the plant calculated from the height, large width and small width to evaluate its regrowth is justified by two main reasons:

Ease of measurement of these dimensions to perform the study of plant regrowth.

The use of these dimensions by authors have conducted similar studies in other regions, ecologically similar to ours, as Euan et al. in the case of the prediction of the components of aboveground biomass of Atriplex canescens L. from the height, and volume of the plant in 1998, or the work of Salah Tag El Din on the prediction of regrowth of Salsola vermiculata L. species phytomass in 1994.

The dimensions of each sampled plant are measured before and after defoliation (corresponding to the defoliation stage and according to the degree of severity of the defoliation), and during the regrowth phase corresponding to the growth stage of the following year. These measures concern the height (H), the large width (LW) and the small width (SW) of each sampled plant to evaluate the initial volume, the removed volume (cut), the final volume and the volume regenerated. Each plant is geometrically assimilate to a truncated cone and the calculation of volumes above - mentioned is determined by the following formulas:

Initial volume : Vi = [3,14x Hi x(LWi2 + LWi x SWi + SWi2)]/3

Vi : initial volume - Hi : initial height - LWi : initial large width - SWi : initial small width.

Removed volume (cut): Vr determined from the initial volume depending on the severity of the cut. Thus, we have three categories of removed volume corresponding to three types of cuts made:

Severe defoliation: Vr = Vi x 0.75

Moderate defoliation: Vr = Vi x 0.50

Light defoliation: Vr = Vi x 0.25.

Final volume: measured during the phenological stage corresponding of the following year

Vf = [3.14 x Hf x (LWf2 + LWf x SWf + SWf2)] / 3 with

Vf: final volume - Hf: final height - GLf: final large width - Plf: final small width.

Regenerated Volume: Vre = Vf - (Vi - Vr) with

Vre: regenerated volume - Vf: final volume - Vi: initial volume - Vr: removed volume.

Regrowth rate of phytomass : calculated from the ratio of the regenerated volume on the removed volume.

RPre = (Vre / Vr) x 100

RPre: regrowth rate - Vre: regenerated volume - Vr: removed volume.

Determination of dry matter

The samples are weighed on field to get the fresh weight of green material. At the laboratory, samples are passed in an oven set at a temperature of 105 ° C for 24 hours until a constant weight. The weigh-out of the oven provides dry matter.

In this study, the dry matter was evaluated for each phenological stage (removed dry matter) and for corresponding phenological regrowth stage (regenerated dry matter). A rate of regenerated dry matter (DMre) is then calculated by expressing the ratio of the regenerated dry matter to the removed dry matter.

DMre (%) = DMre / DMr with

DMre (%): Percentage of dry matter regenerated - DMre: Amount of dry matter regenerated;

DMr: Amount of dry matter removed.

Statistical analysis

STATISTICA software (StatSoft France (2003) STATISTICA (data analysis software), version 6 was used to:

Analysis of variance and the establishment of curves to determine the significance of the effect of defoliation and phenological stage on the rate of the regrowth of phytomass and the rate of dry matter regenerated ; and

The development of mathematical models equations governing the regrowth rate of the phytomass and regenerated dry matter in the case of Darmoune rangeland. This research is based on the equations of linear regression method and the establishment of correlations between the parameters evaluated, namely the regrowth rates of phytomass and dry matter on one hand, and the factors represented by the measured plant dimensions H (height), LW (Large width) and SW (small width) on the other hand.

The search for predictive mathematical equations models based on factors easily quantifiable on terrain allows the development of important tools in the decision support on the management and development of natural steppe rangelands (Al Hartani E and Fogel 1998, Bartolome 2006, Ferchichi 2004, Richardsonet al. 2007, Salah Tag El Din 1994).

This research was based on the use of the least squares method and the significance of the correlation expressed by R2 in the target to prove the validity of the use of the cut dimensions of the plant for the prediction of the desired parameter (Gerry and Michael 2002, Roger et al. 2004, Salah Tag El Din 1994).


Results

Effect of defoliation degree and phenological stage on the regrowth rate of phytomass

The importance of the interaction between the defoliation effect and the effect of phenological stage lies in the balance between optimal grazing intensity to apply to rangeland and choosing the best time to use it. The observed effects of this interaction, illustrated in Figure n°1, shows that the best results were registered by the light defoliation for all phenological stages with a preponderance of early growth stage (EG). The recorded values ​​indicate that the light defoliation made at early growth (EG) allows a good recovery of the productive potential of the rangeland which can exceed the rate of 100%. This implies that livestock can remain at its starting level which would maintain farming activity without negative impact on the course, while the other two types of grazing intensity require a judicious carrying capacity calculation to avoid negative impacts during its usage after regrowth.

Regrowth rates recorded for the late growth stage are far weaker for the three grazing intensities and reveal the danger facing this type of rangeland in case of consecutive exploitation. Results on the full growing stage (FG) show that without a rigorous calculation of the carrying capacity, other consecutive operations may be at high risk for the future of the rangeland, especially for severe grazing intensity.

Figure 1. Effect of the phenological stage and cutting type interaction on the regrowth rate of the phytomass.

 Effect of defoliation type and phenological stage on the regrowth rate of dry matter

The dry matter is important in the study of dynamic behavior of a rangeland because it is the basis of the evaluation of its feeding value and its pastoral value. Our study showed that regrowth of dry matter follows the same logic as that of phytomass, but in significantly different proportions.

The results represented in Figure 2, show that the most significant effect on the dry matter regrowth rate is related to the light cut for all phenological stages which confirms that the greater the number of buds on the plant is kept, the higher is important regrowth.

Figure 2. Effect of the phenological stage and defoliation degree interaction on the regrowth rate of the dry matter
Mathematical equation models of regrowth rate of phytomass and dry matter

The desired model in our study aimed to quantify the rate of regrowth of plant biomass and dry matter content of Darmoune rangeland for each type of cut (severity of grazing).

The regrowth rate of phytomass (RPre) is evaluated based on the values ​​of the initial height (Hi), the large initial width (LWi) and small initial width (SWi) (Tag El Din Salah S., 1994). The same criteria are used for the regenerated dry matter rate (DMre).

Phytomass regrowth rate (RPre)

Regression analysis revealed the following results:

The regrowth rate of phytomass for severe defoliation is a neperian logarithm function of the initial size of the plant:

RPre (%) = 10.177 + 0.169Ln(Hi) + 0.740Ln(LWi) + 0.13 Ln(SWi) . R2 = 0.61 (p=0,01).

The regrowth rate of phytomass for moderate cut is translated by a quadratic function:

RPre (%) = 8.493 + 0.333(Hi)2 + 0.302(LWi)2 + 0.312(SWi)2. R2 = 0.92 (p=0,001).

In the case of the light cut, the rate of regrowth is expressed by a neperian logarithm function:

RPre (%) = 11.7667 + 0.345Ln(Hi) + 0.300Ln(LWi) + 0.306Ln(SWi). R2 = 0.93 (p=0,005).

For all cuts the rate of regrowth of the plant biomass is also a neperian logarithm function of the cut dimensions of the plant:

RPre (%) = 15.349 + 0.527Ln(Hi) + 0.385Ln(LWi)+0.03 Ln(SWi) . R2 = 0.76 (p=0,005).

Dry matter regrowth rate (DMre)

In our study the rate of dry matter regenerated (DMre) is as follows:

A decimal logarithm function of the initial dimensions of the plant for severe defoliation:

DMre (%) = 112.16 + 0.84 log (Hi) - 0.56 log (LWi) + 0.21 log (SWi). R2 = 0.90 (p = 0.001).

For moderate defoliation regrowth follows quadratic function of the initial dimensions of the plant:

DMre (%) = 77.84 + 0.84 (Hi)2 + 0.14 (LWi)2 + 0.77 (SWi)2. R2 = 0.91 (p = 0.001).

As in the case of severe cuts, the regrowth rate for the light cut is expressed by a decimal logarithmic function:

DMre (%) = 94.03 - 3.83 log(Hi) + 0.23 log(LWi) - 4.5 log(SWi). R2 = 0.78 (p = 0.001).

For all cuts, the rate follows a hyperbolic function of the dimensions of the cut:

DMre (%) = 89.78 to 0.89 (1/Hi) + 0.74 (1/LWi) + 0.38 (1/SWi). R2 = 0.90 (p = 0.001).


Discussion

The combined effect of the severity of the cut and the phenological stage shows the interest that should be brought to the choice of the grazing period and stocking rate to set up during the exploitation of the Salsola vermiculata L. steppe rangelands. Indeed, the major difficulty in the management of steppe rangelands lies in the balance between meeting the animal feeding needs and the ability to maintain the productive potential of the rangeland.

Genevieve and Kidney (1995) states that: "the edible portion of the vegetation is the result of irregular and localized weather events." For the same authors, it would be very difficult to ensure that a given area of a rangeland could provide the same amount of fodder at a given time of the year or for subsequent years.

This trend of rangeland variability characteristic in arid and semi arid areas is confirmed by Bartolome (2006). Following the same author the rangeland spatial variability in small scale coupled to with the time dependence of the site, makes very unpredictable the response of the environment system to the management.

The results of our study show a marked variability of recovery of the productive potential of Salsola vermiculata L. steppe rangelands for exploitation in different phases during the year, expressed here by the phenological stages of the plant, which, coupled to varying degrees of severity of grazing, cannot match the production initially consumed by the animal.

Only a light exploitation (25% of total volume of the plant) seems able to allow regrowth close to the volume harvested plant phytomass or dry matter. This is confirmed by Yousfi and Azzouzi (2014) in the study that examined the use of an animal charge 5têtes / ha on a steppe of Artemisia herba alba Asso. rangeland located in the same region as Salsola vermiculata L. rangeland subject of this study.

It follows that the repeated use of these types of rangeland may jeopardize the very existence of their pastoral resources that are weakened by the very sparse nature of their bioclimate, especially in terms of rainfall. This situation is reflected by the state of degradation characterizing the rangelands open to the exploitation throughout the year.

On the other hand, the results obtained for the regrowth rate equations of the phytomass (RPre), in the case of this study show a strong correlation with the selected variable factors for determining the model for all types of cuts, in this case the initial height (Hi), the large initial width (LWi) and small initial width (SWi) This is also indicated by the results obtained by Salah Tag El Din (1994) in his study on the Salsola vermiculata L. species shown below :

The regrowth rate of phytomass for severe defoliation is a decimal logarithm function of the initial size of the plant:

Yc = -1,30 + 39,30logX1 + 33,01logX2 – 27,7logX3

The regrowth rate of phytomass for moderate cut is translated by a quadratic function:

Yc = -66,5 X12 + 2,53X22 – 14,8 X32

In the case of the light cut, the rate of regrowth is also expressed by a quadratic function:

Yc = -96,30 +3,1 X12 +2,64 X22 + 28,6 X32

For all cuts the rate of regrowth of the plant biomass is a linear function of the cut dimensions of the plant:

Yc = 14,4 + 2,1X1 + 1,1X2 – 2,45X3

Yc : current year production - X1 : Height.- X2 : Large width.- X3 : Small width.

It should be noted that the work of Salah Tag El Din was performed on samples of Salsola vermiculata L. plants produced in nurseries and then transplanted on rangeland, which gives them more homogeneity than those used in our study which are indigenous and heterogeneous in their ages and sizes.

However, both studies show that for all cuts, the rate of regrowth of the phytomass can be quantified according to a linear function of the dimensions of the plant.

The relative ease of implementation of these measures on the ground would equip managers and users of Salsola vermiculata L. species steppe rangelands, under similar conditions of the study, by simple tools to predict future production and arrangements during the next season to avoid any risk of loss of natural forage resources, fundamental basis of extensive sheep farming adapted to the conditions of these rangelands.

The resulting equations for the rate of regrowth of the dry matter (DMre) also show a high correlation with the defoliation type variable factors which would lead to conclude that the use of these dimensions would be an appropriate tool in the decision management of such steppe rangelands in similar conditions to the study.

The absence of other results does not allow comparisons to be made and leads to the conclusion that the results of our study cannot be considered as definitive models.


Conclusion


Acknowledgement

This work could not be done without the help of staff and workers of the High Commission for the Development of the Steppe (HCDS), including those of the Regional Office East (RC East Tébessa); we express them our sincere thanks and gratitude for the moral and material support they have shown to us throughout the period for the fulfillment of this modest study.


References

Al Hartani E and Fogel M 1998 A simulation model for evaluating long term impacts of grazing practices. Drylands : sustainable use of rangelands into twenty first century. IFAD : 139-150.

Bartolome J W 2006 Component of managed grazing and complexity of conducting grazing studies.CAL – PAC Society for Range Management Symposium – Grazing for biological conservation: 1-4.

Batabyal A A and Godfrey E B 2002 Rangeland management under uncertainty : A conceptual approach. Journal of Range Management 55: 12-5.

Ferchichi A 2004  La gestion et l’aménagement des parcours en milieux arides : des systèmes locaux de connaissance à la modélisation mathématique. IAMZ-CIHEAM Cahiers Options Méditerranéennes 62: 381-393.

Généviève C G and Rognon P 1995 Désertification et aménagement en Maghreb, MED-CALTUS, 316 p.

Heitschmidt R K and Walker J W 1996 Grazing management technology for sustaining rangeland ecosystems. Rangeland Journal 18 (2): 194-215.

Nedjraoui D 2004 Evaluation des ressources pastorales des régions steppiques algériennes et définitions des indicateurs de dégradation. IAMZ-CIHEAM Cahiers Options Méditerranéennes 62: 239-243.

Nefzaoui A and El Mourid M 2008 Rangeland improvement and management in Arid and Semi-arid environment of West Asia and North Africa. Karia.net : 28p.

Quinn G P and Keough  M J 2002 Experimental design and data analysis for biologists. Cambridge University Press. The Edinburgh Building, Cambridge CB2 2RU, United Kingdom, 557 p.

Richardson F D Hahn B D and Hoffman M T 2007 Modelling the sustainability and productivity of pastoral systems in the communal areas of Namaqualand. Journal of Arid Environment 70: 701-717.

Roger Stern, Richard Coe, Eleanor Allan and Ian Dale 2004 Good statistical practice for natural resources research. CABI publishing. CAB International Wallingford Oxfordshire OX10 8DE UK, 409 p.

Salah S Tag El Din 1994 Prediction of Salsola vermiculata foliage production for different grazing intensities. Journal of King Saud University 6 Agri. Sci. (2): 311-320.

Squires V R 1998 Experts systems and the computer – based decision support system for managing marginal rangeland.Drylands: sustainable use of rangelands into twenty first century. IFAD: 125-136.

Thomson E F, Mirza Sarwat N and Afzal Javed 1998 Predicting the components of aerial biomass of fourwing saltbush from shrub height and volume. Journal of Range Management 51: 323-325.

Westboy M, Walker B and Noy-Meir I 1989 Opportunistic management for rangeland not at equilibrium. Journal of Range Management42 : 265-273.

Wood B 2004 Natural ressource models in the rangelands. A review undertaken for the National Land and Water Ressources (Audit). Australia : 148p. + appendix.

Yousfi M et Azzouzi B 2014 Essai de détermination d’un modèle de régénération d’un parcours steppique à base d’Artemisia herba alba Asso. sous l’effet d’une charge animale de 5 têtes/ha. Livestock Research for Rural Development Volume 26, Article #133. Retrieved July 8, 2014, from http://www.lrrd.org/lrrd26/7/yous26133.html.


Received 5 November 2014; Accepted 30 November 2014; Published 1 January 2015

Go to top