Primary productivity in relation to environmental variables of a semi-arid grassland ecosystem in Kenya

Abstract

A semi-arid grassland was investigated for biomass production and the factors that influence its productivity. Live aboveground biomass ranged from 73 g m-2 in October 1984 to 338 g m-2 in April 1985. Peak live biomass was attained during the rainy season. Peak dead aboveground biomass was attained during the dry season. Dead aboveground biomass ranged from 66 g m-2 in December 1984 to 651 g m-2 in September 1985. From September 1985 dead aboveground biomass kept a constant trend fluctuating between 400 and 600 g m-2. Standing dead biomass ranged betwee? 19 and 457 g m-2 while litter biomass ranged,b,etween 50 and 190 g m-2. Stratified clipping revealed that most of the live aboveground, biomass (64%) occured in the first 0-20 cm canopy layer above the soil surface. Individually, T. triandra contributed 35% of total aboveground live biomass, P. mezianum 26%, other grasses combined - 21% and dicots with sedges 18%. Belowground live biomass was highest during . the dry season and ranged between 60 and 260 g m-2. Dead belowground biomass ranged between 12 and 345 '.:";- . Relative decomposition rate ranged between 0.023 ~nd 0.18 g g-l mon-1 for aboveground dead herbage and 0.076 and 0.335 g g-l mon-1 for belowground dead. Net primary productivity was 1332.4 g m-2 yr-1 (3.65 g m-2 d-1) for aboveground compartment while it was 965.8 g m-2 y r+! (2.65 g m-2 d+ v ) for belowground compartment. Monthly net primary production ranged between 9 and 324 g m-2 for aboveground material and between 8 and 357 g m-2 for belowground material. Turnover rates for different plant materials ranged between 0.4 for aboveground dead and 2.5 for the litter. Leaf area index ranged from 0 to 3.09 while stem and sheath indices ranged from 0 to 0.95 and 0 to 1.53 respectively. Total area index ranged between 0 and 5.57. Total solar radiation averaged 19.7 MJ m-2 d-1 at the top of plant canopy. 64% (or 12 MJ m-2 d-1) was intercepted by the plant canopy during the growing season. The efficiency with which the intercep- , ted energy was used to produce newdpY matter ranged from 0 to 0.31 ~ MJ-l. ~... . Soil moisture content ranged f~om 6% (during the - dry season) to 35% (during the wet season) in the first o - 5 cm of the soil column. Rainfall amounts ranged from o to 212.6 mm. Field measurements of physiological variables showed that seasonal rates of photosynthesis ranged ... from 0 to 26.8 Vmol m-2 S-1 in T. triandra and 0 to 27.1 pmol m-2 S-1 in P. mezianum at midday. At the same time seasonal rates of transpiration ranged from 0.83 to 9.51 mmol m-2 S-1 in T. trindra and 0.93 to 16.2 mmol m-2 S-1 in P. mezianum. Stomatal conductance ranged from 0.095 to 0.533 cm S-1 in T. triandra and 0.148 to 1.1 cm S-1 in P. mezianum. Leaf water potential ranged from -2 MPa to <-4 MPa at midday. Air temperatures were optimum and ranged between 30 and J50C at midday while vapour pressure deficit ranged from 30 to 40 mbars. Top leaves in the plant canopy possessed the highest ra~~s of photo~ynthesis (22 - 27 pmol m-2 S-1) followed by the middle canopy leaves (10 - 12 pmol m-2 , S-1) and finally by the bottom canopy leaves (8 - 10 pmol m-2 s-1).Diurnal course in photosynthetic rates showed that the highasL values were ~ttained between 11.00 and 12.00 hours during the growing season. During the early dry season peak values occured early in the morning hours (10.00 hours) and eveniq$ hours (16.00 hours) with a midday stomatal closure. The same trend of onepeaked trend during the growing season and two-peaked trend during the qEY season was observed for transpiration rate and stomatal conductance. Laboratory work was carried out on stomatal count, chlorophyll determination and photosynthesis by different plant structures (leaf blades, stems, sheaths and inflorescences). Stomatal count revealed that all structures contained stomata which ranged between 836 cm-2 in stems of T. triandra to 29866 cm-2 on the adaxial surface of leaf blade of T. triandra. Total chlorophyll ranged from 0.18 mg g-l fresh weight in the stem of R. repens to 2.49 in the leaf blade of T. triandra. Ratio of chlorophyll a/b was lowest (2.55) in the sheath of C. caesius and highest (5.73) in the leaf blade of P. mezianum. Mean photosynthetic rates ranged from 0.38 pmol m-2 S-1 in the inflorescence of C. caesius to 28.68 pmol m-Z S-1 in the le&f blade of ,T. triandra. Water stress experiment showed that photosynthesis decreased with water stress effects. Mean rates of C02 assimilation decieased by 47% in T. triandra. The decrease-was by 100% in P. mezianum. With rewatering, photosynthetic rates increased by 49% in T. triandra and by 109% in P. mezianum. Both leaf water potential and chl~rophyll levels decreased with increase in water stress. Leaf water potential decreased from between -1.08 and -1.48 MPa to between -1.6 and <-4 MPa~while total chlorophyll decreased from between 1.14 and 2.52 mg g-1 to between 0.99 and 1.76 mg g-1. However, after rewatering leaf water potential increased to between -1.39 and 3.24 MPa while total chlorophyll increased to between 1.05 and 2.1 mg g-l. Leaf anatomy of the grass species used in the water stress experiment revealed that the grass species were either NADP-me or PEP-ck C4 Kranz sub-types. NADP-me species included T.triandra, P. mezianum, C. caesius and C. ciliaris while PEP-ck species were R. repens, E. paspaloides and D. aegyptium. Scanning electron micrographs of two grasses; T. triandra and P. mezianum confirmed that stomata occured on both adaxial and abaxial surfaces of leaf blade. Stomata were often found in crypts in T. triandra. Except in the stem, epidermal cells of T. triandra possessed protrusions called papillae. Epidermal surfaces of leaf blade, stem, sheath and inflorescence possessed numerous ", stomata, prickles, microhairs and waxy cuticle. Spectral reflectance ratio (SRR) data revealed that t.here wa s a good correla!-ion (r = > -0.8) between SRR data and live aboveground biomass during the growing season and that the method was a good predictor of live aboveground biomass. However, during the dry season when plant leaves were senescent, the correlation was 'poor (r = <-0.5) and SRR was not a good method to p~~dict the amount of live aboveground plant biomass from the SRR data .