Trade-offs and synergies between ecosystem productivity and stability in temperate grasslands

Aim

It is crucial to monitor how the productivity of grasslands varies with its temporal stability for management of these ecosystems. However, identifying the direction of the productivity–stability relationship remains challenging because ecological stability has multiple components that can display neutral, positive or negative covariations. Furthermore, evidence suggests that the direction of the productivity–stability relationship depends on the biotic interactions and abiotic conditions that underlie ecosystem productivity and stability. We decipher the relationships between grassland productivity and two components of its stability in four habitat types with contrasting environments and flora.

Location

France.

Time period

2000–2020.

Major taxa

Grassland plant species.

Methods

We used c. 20,000 vegetation plots spread across French permanent grasslands and remotely sensed vegetation indices to quantify grassland productivity and temporal stability. We decomposed stability into constancy (i.e., temporal invariability) and resistance (i.e., maximum deviation from average) and deciphered the direct and indirect effects of abiotic (namely growing season length and nitrogen input) and biotic (namely plant taxonomic diversity, trait diversity and community-weighted mean traits) factors on productivity–stability relationships using structural equation models.

Results

We found a positive relationship between productivity and constancy and a negative relationship between productivity and resistance in all habitats. Abiotic factors had stronger effects on productivity and stability compared with biotic factors. A longer growing season enhanced grassland productivity and constancy. Nitrogen input had positive and negative effects on grassland productivity and resistance, respectively. Trait values affected the constancy and resistance of grassland more than taxonomic and trait diversity, with effects varying from one habitat to another. Productivity was not related to any biotic factor.

Main conclusions

Our findings reveal how vital it is to consider both the multiple components of stability and the interaction between environment and biodiversity to gain an understanding of the relationships between productivity and stability in real-world ecosystems, which is a crucial step for sustainable grassland management.     

A general suppressor of RNA polymerase I,II and III mutations in Saccharomyces cerevisiae

The pem locus, which is responsible for the stable maintenance of the low copy number plasmid R100, contains the pemK gene, whose product has been shown to be a growth inhibitor. Here, we attempted to isolate mutants which became tolerant to transient induction of the PemK protein. We obtained 20 mutants (here called pkt for PemK tolerance), of which 9 were temperature sensitive for growth. We analyzed the nine mutants genetically and found that they could be classified into three complementation groups, pktA, pktB and pktC, which corresponded to three genes, ileS, gltX and asnS, encoding isoleucyl-, glutamyl- and asparaginyl-tRNA synthetases, respectively. Since these aminoacyl-tRNA synthetase mutants did not produce the PemK protein upon induction at the restrictive temperature, these mutants could be isolated because they behaved as if they were tolerant to the PemK protein. The procedure is therefore useful for isolating temperature-sensitive mutants of aminoacyl-tRNA synthetases.     

Organization of the genes necessary for hydrogenase expression in Rhodobacter capsulatus. Sequence analysis and identification of two hyp regulatory mutants

A 25kbp DNA fragment from the chromosome of Rhodobacter capsulatus B10 carrying hydrogenase (hup) determinants was completely sequenced. Coding regions corresponding to 20 open reading frames were identified. The R. capsulatus hydrogenase-specific gene (hup and hyp) products bear significant structural identity to hydrogenase gene products from Escherichia coli (13), from Rhizobium liguminosarum (16), from Azotobacter vinelandii (10) and from Alcaligenes eutrophus (11). The sequential arrangement of the R. capsulatus genes is: hupR2-hupU-hypF -hupS-hupL-hupM-hupD -hupF -hupG -hupH -huoJ -hupK -hypA-hypB-hupR1-hypC -hypD -hypE -ORF19 -ORF20 , all contiguous and transcribed from the same DNA strand. The last two potential genes do not encode products that are related to identified hydrogenase-specific gene products in other species. The sequence of the 12 R. capsulatus genes underlined above is presented. The mutation site in two of the Hup^? mutants used in this study, RS13 and RCC12, was identified in the hypF gene (deletion of one G) and in the hypD qene (deletion of 54 bp), respectively. The hypF gene product shares 45% identity with the product of hydA from E. coli and the product of hypF from R. leguminosarum. Those products present at their N-terminus a Cys arrangement typical of zinc-finger proteins. The G deletion in the C-terminal region of hypF in the RS13 mutant     

INTRACELLULAR AND EXTRACELLULAR AMINO ACID POOLS OF THE MARINE DIATOM PHAEODACTYLUM TRICORNUTUM (BACILLARIOPHYCEAE) GROWN ON UNENRICHED SEAWATER IN HIGH-CELL-DENSITY DIALYSIS CULTURE1

The consumption of inorganic macronutrients (NO₃^?+ NO₂^?, NH₄⁺, and PO₄^?3) and the composition of intra- and extracellular dissolved free amino acid pools (IDFAA and EDFAA, respectively) were determined in continuous-reservoir batch dialysis cultures of the marine diatom Phaeodactylum tricornutum Bohlin maintained on unenriched natural seawater as a growth medium. Nutrient diffusion (Nd), which equals the nutrient uptake of the culture, increased with the cell density and the age of the culture. A concentration of 6.77 × 10⁷ cells · mL^?1 was obtained in stationary phase, which coincided with the NO₃^?+ NO₂^? diffusion limit (Nd_max) of the dialysis apparatus. The Nd_max for NH₄⁺ occurred much earlier, at the end of exponential growth, whereas Nd_max for PO₄^?3 was not attained during the growth cycle of the culture, even in early stationary phase. A significant depletion (77%) of the IDFAA pool during exponential phase was followed by a reestablishment–to approximately 60% of the initial level–of internal pools during linear and stationary growth phases. This recovery occurred during the illuminated portion of the photoperiod (12:12 h LD) and involved principally the amino acids GLN, GLU, β-GLU, and ASN. The recovery of GLN and ASN levels was particularly significant, because the intracellular concentrations of these amino acids were higher at the end of the growth cycle than before. The EDFAA pool was generally dominated by the amino acids SER and GLY+THR; however, during active growth, ORN and LYS often constituted an important fraction. The EDFAA concentration increased until linear growth phase was reached, during which a higher concentration of total free amino acids was attained in darkness than under illumination. The EDFAA component diminished afterward, and in stationary phase this fraction returned to concentrations equivalent to those observed at the beginning of the growth cycle. The variations in EDFAA concentrations were expressed by a pronounced decrease in the cellular excretion of amino acids with increasing cell density. These cellular responses of Phaeodactylum tricornutum in dense culture, specifically the regulation of amino acid excretion and intracellular pool size, may affect the N-conversion coefficient (Y_N). Consequently, by prolonging the linear phase of growth and reducing the concentration of autoinhibitory metabolites by diffusion, a markedly enhanced final cell density can be achieved in cultures grown on natural unenriched seawater.     

Actin and spectrin-like (Mr= 260–240 000) proteins in gregarines

In Gregarina blaberae a M_r = 47 000 and a M_r = 260–240 000 doublet polypeptides reacted in immunoblotting: i) with a polyclonal monospecific rabbit antibody to frog muscular actin, a monoclonal anti-actin antibody against chicken gizzard; and ii) with polyclonal and monoclonal antibodies to human erythrocyte β-spectrin, respectively. The M_r = 47 000 actin-like protein is associated with the ghost and a contractille cytoplasmic extract. The presence of an actin-like protein in Gregarina and Lecudina and its cellular distribution in the cortex indicated that the gliding movement might involve an actin-myosin system in contrast to previous studies. Immunofluorescence showed clear differences between the anterior part of Gregarina and Lecudina which illustrated the high cell polarity of these protozoa. The M_r = 260–240 000 doublet was detected in SDS-PAGE from G. blaberae trophozoite ghosts but not in the cytoplasmic extracts or in extracts from sexual stages, indicating that the presence of these spectrin-like proteins is stage-dependent. Visualization of the M_r = 260–240 000 by immunofluorescence showed clear species differences, with rings arranged perpendicular to the longitudinal narrow folds of G. blaberae, with longitudinal lines underlying the folds of L. pellucida and with lines separating the large folds of Selenidium pendula. The cellular distribution is consistent with a stabilizer function of the spectrin-like proteins in the scaffolding of the cortex of gregarines according to the high diversity of the cell-shape and the cell motility systems in gregarines. The presence of spectrin-like proteins in protozoa and particularly in parasites from primitive arthropods indicated that ancestral spectrin genes could the M_r = 260–240 000 form.     

Nitric oxide does not promote iron release from ferritin

It has been previously reported that iron release from ferritin could be promoted by nitric oxide (NO) generated from sodium nitroprusside. It was thus proposed that some of the toxic effects of NO could be related to its ability to increase intracellular free iron concentrations and generate an oxidative stress. On the contrary, the iron exchange experiments reported here show that NO from S-nitrosothiols is unable to promote iron release from ferritin. The discrepancy may be explained by the disregarded ability of ferrozine, the ferrous trap used in the previous report, to mobilize iron both from ferritin and from sodium nitroprusside spontaneously.     

Isolation and preliminary characterization of gas1-1, a mutation causing partial suppression of the phenotype conferred by the gibberellin-insensitive (gai) mutation in Arabidopsis thaliana (L.) Heyhn

The semi-dominant gai mutation of arabidopsis confers a dark-green dwarf phenotype resembling that of gibberellin (GA)-deficient mutants. In contrast to GA-deficient mutants, gai mutants do not respond to GA treatments and accumulate higher levels of bioactive GAs than are found in wild-type controls. The gai mutation thus alters the responses of plant cells to GA, indicating that the GAI (wild-type) gene product is involved in GA reception and/or signal transduction. Here we describe the isolation and preliminary characterization of a mutation, gas1-1, which is not linked to gai and which partially suppresses the effect of the gai mutation. Double mutant, gai gas1-1, homozygotes are less severely dwarfed and lighter green than gai GAS1 controls. However, comparisons of the effects of treatments with exogenous GA demonstrate that gas1-1 does not increase the GA responsiveness of the gai mutant. Thus the gas1-1 mutation appears to reduce the GA-dependency of plant growth, and identifies a gene (GAS1) whose product is a candidate GA signal-transduction component.Abbreviations GA gibberellin - GA₃ gibberellic acid We thank Maarten Koornneef (Wageningen Agricultural University, The Netherlands) for providing mutant seed stocks; Mark Aarts and Bernard Mulligan (University of Nottingham, UK) for performing the -irradiation. This work was made possible by AFRC/BBSRC PMB Grants PG208/520 and PG208/0600, and by a grant from the Gatsby Charitable Foundation. P.C. was supported by a Human Capital and Mobility Fellowship from the EC.