Journal of Young Investigators
    Undergraduate, Peer-Reviewed Science Journal		 Volume Seven
    RESEARCH ARTICLE
RECENT ISSUES | ARCHIVES | RESOURCES | JYI NEWS | ABOUT JYI 
Issue 1, March 2003

Biological & Biomedical Sciences
Nutrient Content of Floodplain Grasses and Ruminant Feces in the Okavango Delta

Timothy Stowe
Hope College
Advisor: Casper Bonyongo, Ph.D.
University of Botswana

Abstract

Increasing tourism and controlled burning are encroaching on natural habitats within the Moremi Game Reserve of northern Botswana. To gain an understanding of the local ecosystem, nutrient content was determined for the dominant species Panicum repens, Cynodon dactylon, and various sedge species in the Okavango Delta region. Calcium (Ca), potassium (K), phosphorus (P) and nitrogen (N) levels were assessed as indicators of forage quality. Fecal nutrient content of large herbivores endemic to the Delta (Syncerus caffer, Damaliscus lunatus, Connochaetes Albojubatus, Equus burchelli) was also determined. Results show multiple nutrient deficiencies in both plant and ruminant species, most notably for N and P. This study, conducted at the end of the rainy season (April 2002), is the first in a series of studies meant to investigate seasonal variations in nutrient levels as a means for establishing criteria for effective ecological management.

Introduction

"The conclusion of scientists and conservationists is therefore virtually unanimous: the only way to save wild species is to maintain them in their original habitats. Considering how rapidly such habitats are shrinking, even that straightforward solution will be a daunting task. Many ecosystems have already been lost, and others seem doomed." -E.O. Wilson

A place of spectacular beauty, the Okavango River Delta of northern Botswana serves as a testimony to the miracles of biodiversity. The Delta's wetland ecosystem is an endangered environment of international significance, providing a habitat for 2,000-3,000 species of plants, more than 164 species of mammals, 65 fish species, 450 birds and a remarkably diverse population of insects, many of which are endemic to the area (Okavango Wildlife Society 1999). Equally important, the Delta's resources provide a means of living for thousands of people by supplying them with precious water, food, shelter and more recently, employment in Botswana's second most profitable sector, eco-tourism (Bonyongo 1999). However, human exploitation of the Delta's resources does not come without cost. As the pressure of increasing tourism and expansion of humans continues to challenge the Delta, steps must be made to ensure that the Delta ecology and vast biodiversity remain intact.

The Okavango Delta is a large, land-locked alluvial fan situated in the northwest portion of Botswana's semi-arid Kalahari basin. The Delta covers a total area of about 22,000 km2, 27% of which is claimed by permanent swamps, 45-55% of which are seasonal flooded grasslands, and the rest of which is low-lying, dry savannah (Bonyongo 1999). Originating in the highlands of Angola, the delta waters work their way down through Namibia and join the Okavango River in the northwest corner of Botswana. After crossing the Gumare Fault in the north, the river extends southeast across the Kunyere and Thamalakane Fault lines, where it fans out and eventually gives rise to the maze of lagoons, channels, and islands that make up the Delta (McCarthy and Ellery 1997). These unique geomorphological characteristics generate a diverse range of habits, which support the array of wildlife occupying the area.

Seasonally, Delta habitats undergo enormous variations in water levels. The seasonal rains and, more importantly, the seasonal floodwaters that trickle down from the Angolan highlands, are the main factors that influence the yearly flooding (Biggs 1979). With peak flooding occurring at the end of the dry season in July and August, the seasonal floodplains swell with water. Receding in September, the floodwaters ebb, reaching their lowest point in January and February. Ninety-seven percent of the annual inflow, ranging from 6,000 to 16,000 m3, is lost to evapotranspiration and ground leaching, leaving only a small fraction left to exit at the Delta's southern fringes (McCarthy 1998).

The rainfall and flooding patterns are two of the main determinants of floodplain soil properties (Blom 1996, Bonyongo 1999). In turn, these soil properties affect the forage quality, a measure of the composition and nutrient content of plant species. Unfortunately, unnatural processes can also influence forage quality. The establishment of ecotourism camps by safari companies within the Delta has been accompanied by controlled burning to creating artificial vegetation patterns. This alters the natural ecology of vegetation zonation and may impinge upon the homeostatic mechanisms of biodiversity within the Delta.

Although it is understood that the Delta's rapidly changing and varied environment is key to the maintenance of its biodiversity, the contributions of endemic flora and fauna have only begun to be explored (Monna 1999). For example, forage quality has been shown to influence the habitat selection of herbivores residing within the Delta (Vallentine 1990). These include many Delta ruminants, such as buffalo (Syncerus caffer), tssessebe (Damaliscus lunatus), wildebeast (Connochaetes Albojubatus), and zebra (Equus burchelli). Unfortunately, few studies have investigated the effect of forage quality on habitat selection. It is therefore important to conduct further research on forage quality and its effects on fauna, as pressures continue to increase on resident species in the Okavango Delta.

Measurements of forage quality and the nutritional status of endemic fauna can be used as predictors of habitat selection. Some of the most popular methods for determination of forage quality include the analysis of macronutrients, including nitrogen (N), phosphorus (P), calcium (Ca), and potassium (K) content in concert with calculations of stem weight/leaf weight ratios (Voeten and Prins 1999). However, nutrient levels have not alone been proven as a generally reliable method for determination of the nutritional status of their respective grazing ruminants (Howery and Pfister 1990). For example, grazers have the ability to select the most nutritious parts of plants (Howery and Pfister 1990), enabling them to select higher quality forage than that which is randomly selected. Grazing ruminants predominantly focus on foraging green leaves concentrated with the aforementioned nutrients in order to maximize nutritional intake. The nutritional status of grazing ruminants can be elucidated using total fecal N content (Nunez-Hernandez 1992; Dorgeloh 1998), along with comparative analyses of fecal P (Grant 1998), Ca, and K. These nutrients affect the ability of nitrogen-fixating bacteria within the guts of ruminants to extract N from forage. Monitoring nutritional status can be used to explore generalized seasonal nutritional variations and changes between habitats and populations of free-roaming ruminants (Irwin et al. 1993).

The purpose of this study was to establish a baseline for monitoring the nutritional status of large ruminants and forage quality of selected habitats within the Okavango Delta region. The subjects selected were the grasses Panicum repens, Cynodon dactylon, Leersia hexandra and Acroceras macrum, and the commonly observed Delta ruminants S. caffer, D. lunatus, C. Albojubatus, and E. burchelli. The diet preferences for the grazing ruminants are not known, so the grass species were selected by examining frequently grazed areas. The goal is to understand how the nutritional status of these species changes both annually and across seasons. It is anticipated that forage quality will be the primary determinant of the herbivore nutritional status.

This study serves as the first in a series to assess seasonal trends in the nutritional status of Delta flora and fauna. It is hoped that this knowledge will contribute to a more efficient conservation effort through greater ecological understanding of seasonal habitat variation.

Methods and Materials

Study Area

The University of Botswana's Harry Oppenheimer Okavango Research Centre (HOORC) field station (19º35'S 19º35'E) was used as a base camp for this study. Within the Moremi Game Reserve (MGR) of the Okavango Delta (approximately 1000 m above sea level, 515 mm annual rainfall), the base camp is situated along the Boro River at the southwestern end of Chief's Island, near the Nxaraga Lagoon (Fig. 1). Within an approximately 40 km radius from camp, numerous floodplains have been identified as areas of high grazing activity. The dominant grass species in the secondary and upper floodplain grazing areas have been identified as P. repens (torpedo grass) (Fig.2) and C. dactylon (Bermuda grass) (Fig.3), while dominant species in the primary floodplain areas were most often sedges, which include L. hexandra (Southern cut grass) and A. macrum (Nile grass) (Fig. 4) (Bonyongo 1999). Rough visual comparisons were made to divide P. repens into short and tall varieties for further analysis.

 
Fig 1. Vegetation zonation in the Okavango Delta and study site (Stainstreet et al 1993)
 
Fig 2. Floodplain areas dominated by Panicum repens showing both lightly grazed (high, a) and heavily grazed (low, b)
 
Fig 3. Floodplain areas dominated by Cynodon dactylon (a) and sedges (b)
 
Fig 4. Floodplain area within study area showing Syncerus caffer, Connochaetes Albojubatus and Equus burchelli grazing simultaneously (a) and a group of Syncerus caffer grazing in Panicum repens (low).
 

Vegetation Sample Collection

Grass samples were collected from four plant communities, classified as commonly grazed, within a floodplain area (SN2) of the MGR. Samples were chosen from areas exhibiting dominant species of P. repens, C. dactylon, L. hexandra, and A. macrum on April 17, 2002, at the end of the summer rainy season. Sample areas were selected by randomly throwing a 50x50 cm iron square onto one of the aforementioned floodplain communities. All accessible vegetation within the boundaries of the square was removed using scissors, placed into a paper bag and labeled. Vegetation height ranged from 5 to 30 cm. The vegetation that remained in each plot after removal was about 1-2 cm high.

About 24 hours after collection, samples were dried at 60°C for 4 hours and then ground to a fine powder with a Cyclotec 1093 grinding mill (FOSS, South Africa).


Fecal Sample Collection

Fresh fecal samples were collected from the same area as the vegetation samples. Wetness and lack of insect damage were used as indicators of freshness within 24 hours. Two samples each from S. caffer, D. lunatus, C. Albojubatus and E. burchelli were collected. It was assumed that all collected fecal samples came from animals that grazed within the range of the study area and that each sample belonged to a different individual. Fecal samples can be collected up to seven days after defecation for nutritional analysis to estimate fecal nitrogen (FN) and fecal phosphorus (FP) (Leite and Stuth 1994). Given the preliminary nature and time restraints of the study, samples were gathered only to assess FN, FP, and macronutrient levels during the current season.

Samples were collected April 17, 2002, in paper bags, labeled and then dried approximately 24 hours after collection at 80°C for 48 hours. The samples were ground to a fine powder using a Cyclotec 1093 grinding mill.


Sample Digestion

An acid digest was performed on grass and fecal samples in order to prepare them for analysis. A Kjedahl selenium catalyst tablet (Merck, Germany) was combined with approximately 0.3 g of either grass or fecal matter and placed in a 120-ml test tube. A total of 6.0 ml of 98% sulphuric acid was added to each tube. Ten blank samples containing only catalyst tablets and acid were also prepared along with a standard feces and grass sample with known chemical properties. The tubes were allowed to incubate at 400°C for 1 hour in order to ensure complete digestion of the samples. After the hour period, the samples were allowed to cool for 15 minutes, after which, 15.0 ml of deionized H2O was added to each tube. After another cooling period of 30 min, the volume of each tube was brought to 100 ml using deionized H2O. Samples were capped and allowed to sit overnight to allow for equilibration of ion content and dissolved elements.


Analysis of Macronutrients

Grass and fecal samples were analysed for Ca and K content using atomic absorption spectroscopy (Varian Inc, USA). Standards of 1, 5, 10, and 20 ppm were prepared for Ca by dilution of a stock solution of 1000 ppm (mg/L). A standard curve was obtained and the Ca ppm was then converted to mg/sample and compared with actual sample weights, in order to obtain percent Ca in sample. The same procedure was followed for analysis of K content, using standards of 1, 5, 10, and 15 ppm. Samples and blanks were diluted 3 ml in 25 ml.


Analysis of Total N and P

Samples were analysed for N and P using a Bran + Luebbe auto analyzer (AA3; Roselle, IL ). Nitrogen standards of 5, 10, 20, 50, and 100 ppm were prepared as described in AutoAnalyzer Applications in order to obtain a concentration curve. N was analyzed by reacting samples with salicylate and sodium hypochlorite solutions in the presence of nitroprusside catalyst to analyze for a blue compound that absorbs at 660 nm (Bran + Luebe 2000). The Bran + Luebbe AA3 was also used to analyze P content in samples. Standards of 2, 4, 8, 12, 20, 40, and 50 ppm were prepared as described in AutoAnalyzer Applications. P in samples was quantified by absorbance of a blue compound formed from the ortho-phosphate products of the acid digest with molybdate, phosphate, and antimony, and ascorbic acid reduction at 660 nm (Bran + Luebe 2000).


Statistical Analysis

Given that the nutritional status of these samples reflects only their condition at the given point in the season, only comparisons between groups and expected baseline levels could be made. An ANOVA analysis of variance was conducted on groups of grass samples and feces samples separately followed by Tukey's test to determine which groups differed significantly. SPSS statistical analysis software (SPSS Inc., Chicago, IL) for Windows was used.

Results

Macronutrient levels

To interpret the data, critical baseline values for the grass species were taken from Jones et al. (1991). These values were established by examination of related species common to similar habitats in South Africa and are comparable among species.

The K content in grass species is considered deficient when it is less than 1.50% and excessive when it is more than 3.00 %. The K values obtained for the tall variety of P. repens (0.988% ± 0.285 %) and the sedges (1.172% ± .076%) were lower than the critical level, whereas those for C. dactylon and the short variety of P. repens were within one standard deviation of the expected range (Table 1, Fig. 5). There was no significant difference between plant species regarding K content.

 
Table 1. Vegetation nutritive values of areas with differing dominant species sampled on March 17 2002. *indicates deficiency
 
Fig. 5. Vegetative chemical components of floodplain grasses collected on March 17, 2002. *indicates deficiency
 

Ca content in vegetation usually ranges from 0.20% - 3.00%, and between 0.30 % and 1.00 % is considered sufficient. The Ca values obtained for the tall variety of P. repens (0.215% ± 0.024 %) and the sedges (0.204% ± 0.015%) were again lower than the critical level, whereas those for C. dactylon and the short variety of P. repens were within one standard deviation of the expected range (Table 1, Fig. 5). A significant difference in Ca content existed between the tall variety of P. repens and the sedges when compared with the short variety of P. repens (p<0.05).

Although macronutrient levels were analysed in ruminant fecal samples, it was only done considering future potential for a seasonal comparison of nutrient levels. Also, these values can be used to gauge nutritive differences between habitats within the Delta (Table 2). Macronutrient values were determined for each species: D. lunatus (Ca 0.636% ± 0.016%, K 0.874% ± 0.017%), S. caffer (Ca 0.686% ± 0.003%, K 1.924% ± 0.243%), C. Albojubatus (Ca 0.625% ± 0.050%, K 1.128% ± 0.237%), and E. burchelli (Ca 0.154% ± 0.042%, K 1.716% ± 0.863%).

 
Table 2. Faecal nutritive values of different species sampled on March 17 2002 *indicates deficiency
 


N and P levels

All vegetation tested displayed N levels well below the required minimum percentage. P. repens (tall variety) was significantly low in N content when compared with P. repens (short variety) and the sedges (p<0.05) (Table 1, Fig. 5).

P is considered deficient when levels are lower than 0.20% and in excess when they are above 1.00%. All vegetation, with the exception of P. repens (tall variety), exhibited P levels lower than the critical level (Table 1, Fig. 5). There was no significant difference among species.

Fecal N values obtained for D. lunatus (0.803% ± 0.159%) and E. burchelli (0.693% ± .404%) were slightly below the baseline for proper rumen fermentation, whereas those S. caffer and C. Albojubatus were within standard deviations of the baseline. No significant difference was observed in fecal N levels between species (Table 2, Fig. 6).

All ruminant species displayed fecal P levels lower than the critical value of 0.2%. There was no significant difference between species in fecal P content.

 
Fig. 6. Fecal chemical components of delta ruminants collected on March 17, 2002. *indicates deficiency
 

 

Discussion

Of the grasses, P. repens (short variety) and C. dactylon displayed the best overall nutritional status, being deficient only in N and P. For comparison, the sedges and P. repens (tall variety) were deficient in all four nutrients, with the exception that the tall variety of P. repens had adequate levels of P. This result is not surprising, as P. repens (short variety) and C. dactylon grow in the most heavily grazed regions of the floodplain. As plants age, they generally decline in nutritive value. Constant grazing pressure stimulates constant regrowth, thereby allowing plants to maintain a "younger" nutritive state through leaf regeneration (Bonyongo 1999).

Interestingly, almost all plant species had insufficient levels of P and N. As N is an integral component of protein and DNA and P is an integral component of protein, this may have interfered with cell division and renewal and reduced plant protein content. In any case, it is important to note that these deficiencies indicate substandard soil nutrient content. In corroboration, P has been identified as the most deficient major element in the soils and plants of Botswana (APRU 1978). It is not known why the tall variety of P. repens had adequate levels of P, but inadequate levels of all other nutrients tested.

For the grazing ruminants, no baseline levels have yet been established for Ca and K. From the data, though, Ca levels appear to be identical for all ruminants except for E. burchelli, which exhibited about one-quarter the fecal Ca content. Since baseline levels are not available for this nutrient, it has only been speculated that this represents a dietary deficiency of this crucial mineral. Interestingly, P levels varied widely among herbivores, which may be explained by varying levels of plant P content and dietary preferences.

In contrast, baseline levels of N and P have been established. Fecal N levels are an indication of dietary protein intake. Ruminants, including wild ungulates, require a minimum of 5% crude protein (0.8% N) in their food to maintain body weight (Liversidge and Berry 1995). Also, proper rumen fermentation is disrupted if fecal N concentrations are not at least 1.10-1.20% (Grant et al. 1995). Both fecal N and P levels were found to be lower than what is acceptable for proper rumen function, as defined in previous studies (Dorgeloh et al. 1998). This coincides with suboptimal plant N and P content, so it is likely to be a direct effect of dietary intake. It should be noted that although fecal macronutrient levels cannot be used to directly gauge the nutrient status ruminants, Irwin et al. (1993) have shown that these values can be used to compare differences in nutrient levels across changing habitats and variations in seasons.

Several factors may have affected the results of this study. First, all samples were collected toward the end of the summer rainy season. The summer rainy season in the Delta is marked by occasional showers taking the form of short local thunderstorms between November and March, sometimes leading into April (Bonyongo 1999). Given that all samples were collected on April 17, 2002, in a year that received less rain than was normally expected, and that seasonal floodwaters had not yet arrived, it is likely that the subjects studied were under considerable stress. Apparently, about once every 10 years, stress induced by water shortage reaches levels similar to that witnessed during the study period (Bonyongo, personal communication). Thus, it is important that this be taken into account when interpreting the data, especially when making seasonal comparisons.

Secondly, small sample size may have also affected the results of this study. As only two samples for each vegetation type and ruminant species were analyzed, it is statistically possible that unrepresentative data were obtained. Future studies should increase sample size.

Given the short duration of this study, the small sample size and the anomalous shortage of rainfall, only incomplete conclusions regarding the current state of the nutritional status of the forage and ruminants studied can be drawn. In order to obtain a better understanding of this perceived stress and concomitant nutritional fluctuations, it is important to continue data collection between seasons.

The Okavango Delta ecology and its unique interactions with the flora and fauna have only recently fallen under scrutiny. Although baseline levels for most macronutrients and elements, such as N and P, have been established, it is possible that the flora and fauna of the Delta have adapted differently than life in other regions of Africa. Therefore, the seasonal aspects of this study and future investigations are important to understanding the natural ecology of the region and how to preserve its biodiversity.

Human disturbances of natural vegetation patterns and habitat fractionation are encroaching upon natural habitats within the Delta. The largest culprit is the growing tourism industry. Safari companies use controlled burning to create artificial plant zonation, to stimulate plant regrowth and to improve forage quality. This in turn attracts large hervibores, which draws more customers. Although burning is a natural maintenance mechanism of the Delta ecology, not enough is known about how artificial burning affects the local ecology. Certainly, if controlled burning continues to be unregulated, the potential for long-term damage to the natural environment is serious It is possible that burning can be safely used to enhance the tourism industry and maximize the efficiency of sustainable management, but more research is needed.


 
Fig. 7. Giraffa camelopardalis seen walking through study area with fire in background (a) and picture of a recently burnt area within the study area (b).
 


As suggested in the quotation that opens this paper, the best way to maintain biodiversity is to preserve its existence in its natural habitat. Considering the swiftness and increasing extent of human-induced environmental damage in the Okavango Delta, it is necessary to understand the ecology of the region. Only with this knowledge can an educated attempt toward conservation of one of the planet's last remaining large wetlands be achieved. 
References

A.P.R.U. 1978. Beef cattle and range research programme in Botswana, 1970-1977. Ministry of Agriculture, Gaborone, Botswana.

Biggs, Russell C. (1979) The ecology of Chief's Island and the adjacent floodplains of the Okavango Delta, Botswana. Masters, University of Pretoria, Pretoria.

Bran+Luebbe AutoAnalyzer Applications. (2000).

Blom, C.W.P.M. and Voesenek,L.A.C.J. (1996). Flooding: the survival strategies of plants. Tree. 11: 290-295

Bonyongo, M. C. (1999). Vegetation ecology of the seasonal floodplains in the Okavango Delta, Botswana. Master Thesis, University of Pretoria, Pretoria.

Bonyongo, M. C. (2002). Personal communication.

Dorgeloh, W.G., Van Hoven, W., and Rethman, N.F.G. (1998). Faecal analysis as an indicator of the nutritional status of the diet of roan antelope in South Africa. S. Afr. Wildl. Res. 28:16-22.

Grant, C.C., Meissner, H.H. and Schultheiss, W.A. (1995). The nutritive value of veld as indicated by faecal phosphorus and nitrogen and its relation to the condition and movement of ruminants during the 1992-1993 drought in the Kruger National Park. Koedoe 38(1): 17-31

Howery, L.D. and Pfister, J.A. (1990). Dietary and faecal concentrations of nitrogen and phosphorus in penned white-tailed deer does. J. Wildl. Manage. 57(2): 413-421.

Irwin, L.L., Cook, J.G., McWhirter, D.E., Smith, S.G. and Arnett, E.B. (1993). Assesing winter dietary quality in bighorn sheep via faecal nitrogen. J. Wildl. Manage. 54(3): 389-391.

Jones, J.B., Wolf, B., and Mills, H.A. 1991. Plant analysis handbook, a practical sampling, analysis, and interpretation guide. Micro-Macro Publsihing, Inc.

Leite, E.R. and Stuth, J.W. 1994. Influence of duration of exposure to field condition on viability of faecal samples for NIRS analysis. J. Range. Manage. 47(4): 312-314.

Liversidge, R. and Berry, M.P.S. 1995. Voeding. In: Bothma, J. du P. (ed). Wildplaasbestuur. Van Schaik, Pretoria.

Monna, S.C. (1999). A Framework for international cooperation for the management of the Okavango Basin and Delta. Ramsar Convention on Wetlands. COP7 DOC.20.5.

McCarthy, T.S., and Ellery, W.N. (1997) The fluvial dynamics of Maunachira channel system: northeastern Okavango swamps, Botswana. Water SA. 23(2) April.

McCarthy, T.S., Bloem, A., Larkin, P.A. (1998) Observations on the hydrology and geohydrology of the Okavango Delta. SA J. Geo. 101: 101-117.

Okavango Wildlife Society. (1999). Delta Project Outline. Garsfontein, Pretoria.

Stanistreet, I.G. and McCarthy, T.S. (1993). The Okavango Fan and classification of subaerial fan system. Sedimentary Geology. 85: 115-133.

Vallentine, John F. (1990). Grazing Management. Academic Press, Inc. San Diego. p. 55-70.

Voeten, M.M. & Prins, H.H.T. (1999). Resource partitioning between sympatric wild and domestic hrbivores in the Tarangire region of Tanzania. Oecologia 120: 287-294.

Journal of Young Investigators. 2003. Volume Seven.
Copyright © 2003 by Timothy Stowe and JYI. All rights reserved.
 		 
SEARCH   |   SITE MAP   |   RECENT WEB SITE ADDITIONS          PRIVACY POLICY  |    CONTACT US

JYI is supported by: The National Science Foundation, The Burroughs Wellcome Fund, Glaxo Wellcome Inc., Science Magazine, Science's Next Wave, Swarthmore College, Duke University, Georgetown University, and many others.
Copyright ©1998-2003 The Journal of Young Investigators, Inc.