A yield architecture framework to explain adaptation of pearl millet to environmental stress

Functional knowledge of the physiological basis of crop adaptation to stress is a prerequisite for exploiting specific adaptation to stress environments in breeding programs. This paper presents an analysis of yield components for pearl millet, to explain the specific adaptation of local landraces to stress environments in Rajasthan, India. Six genotypes, ranging from high-tillering traditional landraces to low-tillering open-pollinated modern cultivars, were grown in 20 experiments, covering a range of nonstress and drought stress patterns. In each experiment, yield components (particle number, grain number, 100 grain mass) were measured separately for main shoots, basal tillers, and nodal tillers. Under optimum conditions, landraces had a significantly lower grain yield than the cultivars, but no significant differences were observed at yield levels around 1 ton ha(-1). This genotype x environment interaction for grain yield was due to a difference in yield strategy, where landraces aimed at minimising the risk of a crop failure under stress conditions, and modem cultivars aimed at maximising yield potential under optimum conditions. A key aspect of the adaptation of landraces was the small size of the main shoot panicle, as it minimised (1) the loss of productive tillers during stem elongation; (2) the delay in anthesis if mid-season drought occurs; and (3) the reduction in panicle productivity of the basal tillers under stress. In addition, a low investment in structural panicle weight, relative to vegetative crop growth rate, promoted the production of nodal tillers, providing a mechanism to compensate for reduced basal tiller productivity if stress occurred around anthesis. A low maximum 100 grain mass also ensured individual grain mass was little affected by environmental conditions. The strategy of the high-tillering landraces carries a yield penalty under optimum conditions, but is expected to minimise the risk of a crop failure, particularly if mid-season drought stress occurs. The yield architecture of low-tillering varieties, by contrast, will be suited to end-of-season drought stress, provided anthesis is early. Application of the above adaptation mechanisms into a breeding program could enable the identification of plant types that match the prevalent stress patterns in the target environments. (C) 2003 E.J. van Oosterom. Published by Elsevier Science B.V. All rights reserved.

Projecto de um sistema de rega para uma exploração agrícola na Beira Gomes

Trabalho de Projecto de Natureza Científica para obtenção do grau de Mestre em Engenharia Civil

Grass-endophyte coevolution and ploidy levels in fescues

ABSTRACT Fescues consist of wild and cultivated grasses that have adapted to a wide range of environmental conditions. They are an excellent model species for evolutionary ecology studies that investigate symbiosis and polyploidization and their effects on plant performance. First, they are frequently infected with symbiotic endophytic fungi known to affect a plant’s ability to cope with biotic and abiotic environmental factors. Second, fescue species have been reported to have substantial intraspecific variation in their ploidy level and morphology. In my thesis, I examined large-scale generalizations for frequency of polyploidy and endophyte infections and their effects on plant morphology. As a model species, I selected red (Festuca rubra) and viviparous sheep’s (F. vivipara) fescues. They are closely related, but they differ in terms of distribution and endophyte infection frequency. I investigated the biogeographic pattern and population biology of 29 red and 12 viviparous sheep’s fescue populations across ≈300 latitudes in Europe (400-690 N). To examine plant ploidy levels, I implemented time- and cost-efficient plate-based high throughput flow cytometric analysis. This efficient procedure enabled me to analyze over 1000 red fescue individuals. I found three ploidy levels among them: overall 84 %, 9 % and 7 % of the red fescue plants were hexaploid, tetraploid and octoploid, respectively. However, all viviparous sheep’s fescue plants were tetraploid. Ploidy level of red fescue appeared to some extent follow gradients in latitude and primary production as suggested by previous studies, but these results could be explained better by taking the sampling design and local adaptation into account. Three Spanish populations were mostly tetraploids and one high elevation population in northernmost Finland (Halti) was octoploid, while most other populations (25 sites) were dominated by hexaploids. Endophyte infection frequencies of wild fescue populations varied from 0 to 81 % in red fescue populations and from 0 to 30 % in viviparous sheep’s fescue populations. No gradients with latitude or primary production of the sites were detected. As taxonomy of red fescues is somewhat unclear, I also studied morphology, ploidy variation and endophyte status of proposed subspecies of European red fescues. Contrary to previous literature, different ploidy levels occurred in the same subspecies. In addition to wild fescues, I also used two agronomically important cultivars of meadow and tall fescue (Schedonorus phoenix and S. pratensis). As grass-legume mixtures have an agronomic advantage over monocultures in meadows, I carried out a mixture/competition experiment with fescues and red clover to find that species composition, nutrient availability and endophyte status together determined the total biomass yield that was higher in mixtures compared to monocultures. The results of this thesis demonstrate the importance of local biotic and abiotic factors such as grazing gradients and habitat types, rather than suggested general global geographical or environmental factors on grass polyploidization or its association with symbiotic endophytic fungi. I conclude that variation in endophyte infection frequencies and ploidy levels of wild fescues support the geographic mosaic theory of coevolution. Historical incidents, e.g., glaciation and present local factors, rather than ploidy or endophyte status, determine fescue morphology.

REDUCING NUISANCE CANADA GOOSE PROBLEMS THROUGH HABITAT MANIPULATION

Urban populations of Canada geese (Branta canadensis) cause considerable problems when large numbers congregate in parks, playing fields, and backyards. In most cases, geese are drawn to these sites to feed on the lawns. I tested whether geese have feeding preferences for different grass species. Captive Canada geese preferred Kentucky bluegrass (Poa pratensis) and disliked tall fescue (Festuca arundinaceae) over colonial bentgrass (Agrostis tenuis cv. Highland), perennial ryegrass (Lolium perenne), and red fescue (Festuca rubra). They refused to eat some other ground covers such as pachysandra (Pachysandra terminalis) and English ivy (Hedera helix). These results suggest that goose numbers at problem sites could be reduced by changing the ground cover. I also compared the characteristics of foraging sites used by geese to other foraging sites that geese avoided. Occupied sites were more open so that geese had clearer visibility and greater ease in taking off and landing. This suggests that goose numbers at problem sites also could be reduced by planting tall trees to make it harder for the geese to fly away, and planting bushes and hedges to obstruct a goose's visibility.

Morphogenesis in guinea grass pastures under rotational grazing strategies

This study was conducted in order to evaluate the morphogenetic and structural characteristics of guinea grass cv. Mombasa under three post-grazing heights (intense - 30 cm, lenient - 50 cm and variable - 50 in spring-summer and 30 cm in autumn-winter) when sward light interception reached 95% during regrowth. Post-grazing heights were allocated to experimental units (0.25 ha) in a completely randomized block design with three replications. Post-grazing heights affected only leaf elongation rate and the number of live leaves. Pastures managed with variable post-grazing height showed higher leaf elongation rate in the summer of 2007. This management strategy also resulted in a higher number of live leaves. During the spring of 2006, plants showed lower leaf elongation rate, leaf appearance rate and number of live leaves, and greater phyllochron and leaf lifespan. In contrast, during the summer of 2007, the leaf appearance rate, leaf elongation rate, number of live leaves, and final leaf length were greater while phyllochron, stem elongation rate, and leaf senescence rate were lower. The management of the guinea grass cv. Mombasa with intense or variable post-grazing height throughout the year seems to represent an interesting management target, in terms of leaf appearance rate and number of live leaves.

Integrating the Use of Spring- and Fall-Calving Beef Cows in a Year-round Grazing System (A Progress Report)

Animal production, hay production and feeding, and the yields and composition of forage from summer and winter grass-legume pastures and winter corn crop residue fields from a year-round grazing system were compared with those of a conventional system. The year-round grazing system utilized 1.67 acres of smooth bromegrass-orchardgrass-birdsfoot trefoil pasture per cow in the summer, and 1.25 acres of stockpiled tall fescue-red clover pasture per cow, 1.25 acres of stockpiled smooth bromegrass-red clover pasture per cow, and 1.25 acres of corn crop residues per cow during winter for spring- and fall-calving cows and stockers. First-cutting hay was harvested from the tall fescue-red clover and smooth bromegrass-red clover pastures to meet supplemental needs of cows and calves during winter. In the conventional system (called the minimal land system), spring-calving cows grazed smooth bromegrass-orchardgrass-birdsfoot trefoil pastures at 3.33 acres/cow during summer with first cutting hay removed from one-half of these acres. This hay was fed to these cows in a drylot during winter. All summer grazing was done by rotational stocking for both systems, and winter grazing of the corn crop residues and stockpiled forages for pregnant spring-calving cows and lactating fall-calving cows in the year-round system was managed by strip-stocking. Hay was fed to springcalving cows in both systems to maintain a mean body condition score of 5 on a 9-point scale, but was fed to fall-calving cows to maintain a mean body condition score of greater than 3. Over winter, fall-calving cows lost more body weight and condition than spring calving cows, but there were no differences in body weight or condition score change between spring-calving cows in either system. Fall- and spring-calving cows in the yearround grazing system required 934 and 1,395 lb. hay dry matter/cow for maintenance during the winter whereas spring-calving cows in drylot required 4,776 lb. hay dry matter/cow. Rebreeding rates were not affected by management system. Average daily gains of spring-born calves did not differ between systems, but were greater than fall calves. Because of differences in land areas for the two systems, weight production of calves per acre of cows in the minimal land system was greater than those of the year-round grazing system, but when the additional weight gains of the stocker cattle were considered, production of total growing animals did not differ between the two systems.

Efficacy of Grazing Stockpiled Perennial Forages for Winter Maintenance of Beef Cows

In a three year study, wintering systems utilizing the grazing of stockpiled perennial hay crop forages or corn crop residues were compared to maintaining cows in a drylot. In the summer of 1992, two cuttings of hay were harvested (June 22 and August 2) from three 10-acre fields containing “Johnstone” endophyte-free tall fescue and “Spreador II” alfalfa, and one cutting of hay was harvested from three 10- acre fields of smooth brome grass. “Arlington” red clover was frost-seeded into the smooth bromegrass fields in 1993 and into tall fescue-alfalfa and smooth bromegrass fields into 1994. Two cuttings of hay were harvested from all fields in subsequent years, and three-year average hay yields for tall fescue-alfalfa and smooth bromegrass-red clover were 4,336 and 3,481 pounds per acre, respectively. Regrowth of the forage following the August hay harvest of each year was accumulated for winter grazing. Following a killing frost in each year, two fields of each stockpiled forage were stocked with cows in midgestation at two acres per cow. Two 10-acre fields of corn crop residues were also stocked at two acres per cow, following the grain harvest. Mean dry matter forage yields at the initiation of grazing were 1,853, 2,173 and 5,797 pounds per acre for fields containing tall fescue-alfalfa, smooth bromegrass-red clover, and cornstalks, respectively. A drylot was stocked with 18 cows in 1992 and 1993 and 10 cows in 1994. All cows were fed hay as necessary to maintain a body condition score of five. During grazing, mean losses of organic matter were -6.4, -7.6, and -10.7 pounds per acre per cow from tall fescue-alfalfa, smooth bromegrass-red clover, and cornstalk fields. Average organic matter loss rates from stockpiled forages due to weathering alone were equal to only 30% of the weathering losses of the corn crop residues. In vitro digestibility of both stockpiled forages and cornstalks decreased at equal rates during grazing each year, with respective annual loss rates of .14, .08, and .06% per day. Cows grazing corn crop residues required an average of 1,321 pounds per cow less hay than cows maintained in the drylot to maintain equivalent body condition during the grazing season. Cows grazing tall fescue-alfalfa or smooth bromegrass-red clover had body weight gains and condition score changes equal to cows maintained in a drylot but required 64% and 62% less harvested hay than cows in the drylot during the grazing season. Over the entire stored forage cows grazing tall fescue-alfalfa and smooth bromegrass-red clover required an average of 2,390 and 2,337 pounds per cow less than those maintained in the drylot. Because less hay was needed to maintain cows grazing stockpiled forages, average annual excesses of 5,629 and 3,868 pounds of hay dry matter per cow remained in the stockpiled tall fescue-alfalfa and smooth bromegrass-red clover systems.

Evaluation of Year-round Forage Management Systems for Spring- and Fall-Calving Beef Cows (A Progress Report)

A year-round grazing system for spring- and fall-calving cows was developed to compare animal production and performance, hay production and feeding, winter forage composition changes, and summer pasture yield and nutrient composition to that from a conventional, or minimal land system. Systems compared forage from smooth bromegrass-orchardgrass-birdsfoot trefoil pastures for both systems in the summer and corn crop residues and stockpiled grass-legume pastures for the year-round system to drylot hay feeding during winter for the minimal land system. The year-round grazing system utilized 1.67 acres of smooth bromegrassorchardgrass- birdsfoot trefoil (SB-O-T) pasture per cow in the summer, compared with 3.33 acres of (SB-O-T) pasture per cow in the control (minimal land) system. In addition to SB-O-T pastures, the year-round grazing system utilized 2.5 acres of tall fescue-red clover (TFRC) and 2.5 acres of smooth bromegrass-red clover (SBRC) per cow for grazing in both mid-summer and winter for fall- and spring-calving cows, respectively. First-cutting hay was harvested from the TF-RC and SB-RC pastures, and regrowth was grazed for approximately 45 days in the summer. These pastures were then fertilized with 40 lbs N/acre and stockpiled for winter grazing. Also utilized during the winter for spring-calving cows in the year-round grazing system were corn crop residue (CCR) pastures at an allowance of 2.5 acres per cow. In the minimal land system, hay was harvested from three-fourths of the area in SB-O-T pastures and stored for feeding in a drylot through the winter. Summer grazing was managed with rotational stocking for both systems, and winter grazing of stockpiled forages and corn crop residues by year-round system cows was managed by strip-stocking. Hay was fed to maintain a body condition score of 5 on a 9 point scale for spring-calving cows in both systems. Hay was supplemented as needed to maintain a body condition score of 3 for fall-calving cows nursing calves through the winter. Although initial condition scores for cows in both systems were different at the initiation of grazing for both winter and summer, there were no significant differences (P > .05) in overall condition score changes throughout both grazing seasons. In year 1, fall-calving cows in the year-round grazing system lost more (P < .05) body weight during winter than spring-calving cows in either system. In year 2, there were no differences seen in weight changes over winter for any group of cows. Average daily gains of fall calves in the yearround system were 1.9 lbs/day compared with weight gains of 2.5 lbs/day for spring calves from both systems. Yearly growing animal production from pastures for both years did not differ between systems when weight gains of stockers that grazed summer pastures in the year-round grazing system were added to weight gains of suckling calves. Carcass characteristics for all calves finished in the feedlot for both systems were similar. There were no significant differences in hay production between systems for year 1; however, amounts of hay needed to maintain cows were 923, 1373, 4732 lbs dry matter/cow for year-round fall-calving, year-round spring-calving, and minimal land spring-calving cows, respectively. In year 2, hay production per acre in the minimal land system was greater (P < .05) than for the year-round system, but the amounts of hay required per cow were 0, 0, and 4720 lbs dry matter/cow for yearround fall-calving, year-round spring-calving, and minimal land spring-calving cows, respectively.

Evaluation of a Year-Round Grazing System: Winter Progress Report

The winter component of a year-round grazing system involving grazing of corn crop residues followed by grazing stockpiled grass-legume forages was compared at the McNay Research Farm with that of the winter component of a minimal land system that maintained cows in drylot. In the summers of 1995 and 1996, two and one cuttings of hay per year were harvested from two 15-acre fields containing “Johnston” low endophtye tall fescue and red clover. Two cuttings of hay in 1995 and one cutting in 1996 were harvested from two 15-acre fields of smooth bromegrass and red clover. Hay yields were 4,236 and 4,600 pounds of dry matter per acre for the tall fescue-red clover in 1995 and 1996, and 2,239 and 2,300 pounds of dry matter per acre for the smooth bromegrass-red clover in 1995 and 1996. Following grain harvest, four 7.5-acre fields containing corn crop residues were stocked with cows at midgestation at an allowance of 1.5 acres per cow. Forage yields at the initiation of corn crop grazing in 1995 and 1996 were 3,757 and 3,551 pounds of dry matter per acre for corn crop residues. Stockpiled forage yields were 1,748 and 2,912 pounds of dry matter for tall fescue-red clover and 1,880 and 2,187 pounds for smooth bromegrass-red clover. Corn crop residues and stockpiled forages were grazed in a strip stocking system. For comparison, 20 cows in 1995 and 16 cows in 1996 were placed in two drylots simultaneously with initiation of corn crop grazing, where they remained throughout the winter and spring grazing periods. Cows maintained in drylots or grazing corn crop residue and stockpiled forages were supplemented with hay as large round bales to maintain a body condition score of five. In both years, no seasonal differences in body weight and body condition score were observed between grazing cows or cows maintained in drylots, but grazing cows required 85% and 98% less harvested hay in years 1 and 2 than cows in drylot during the winter and spring. Because less hay was needed to maintain grazing cows, excesses of 12,354 and 5,244 pounds of hay dry matter per cow in 1995 and 1996 remained in the year-round grazing system. During corn crop grazing, organic matter yield decreased at 23.5 and 28.8 pounds of organic matter per day from grazed areas of corn crop residues in 1995 and 1996. Organic matter losses due to weathering were 6.8, 10.3, and 12.7 pounds per day in corn crop residue, tall fescue-red clover and smooth bromegrass-red clover in 1995 and 12.1, 10.7, and 12.1 in 1996. Organic matter losses from grazed and ungrazed areas of tall fescue-red clover and smooth bromegrass-red clover during stockpiled grazing were 6.9, 6.9, and 2.1, 2.9 in 1995 and 13.4, 4.3, and +6.9, 4.4 pounds per day in 1996.

Evaluation of a Year-Round Grazing System: Summer Cow-Calf Progress Report

A comparison was made between two different summer grazing systems at the McNay Research Farm. One system was the summer component of a year-round grazing system, involving the rotational stocking of smooth bromegrass-orchardgrass-birdsfoot trefoil pastures and winter stockpile pastures with cow-calf pairs co-grazing with stocker yearlings at .75 animal units per acre. That system was compared with a minimal land system involving the rotational stocking of smooth bromegrass-orchardgrass-birdsfoot trefoil summer pastures with cow-calf pairs grazing at .64 animal units per acre and hay removal from 25% of the pasture. Stocker yearlings or hay removal were used as management tools to remove excess forage and optimize forage quality. Hay was removed once from three fourths of the winter stockpiled pastures in 1996 (Yr. 1) and all the pasture in 1997 (Yr. 2). One hay removal occurred on one fourth of the allocated summer pastures in Year 1 and one half of the pastures in Year 2. In Year one, cow-calf pairs grazing in the year-round system utilized one fourth of the winter stockpile pastures due to a lack of forage on the summer pastures, whereas in Year 2 cowcalf pairs grazed winter stockpile pastures to remove forage as a second cutting of hay. Cow-calf pairs grazing with hay removal were supplemented with harvested hay for two weeks during the summer of Year 1 due to lack of grazable forage; in Year 2, no supplementation was needed. Grazing system did not affect cow body weight, condition score, or daily calf gain in either year. Growing animal production per acre was affected by grazing system, with the minimal land system having a higher production level in Year 1 and Year 2. The year-round system also produced more net winter forage than did the minimal land system in Year 1. Differences in forage yield and quality were only observed between winter stockpile forages of tall fescue-red clover and smooth bromegrass-red clover and summer pastures during the months of June, July, and August.

Evaluation of a Year-Round Grazing System: Winter Progress Report

The winter component of a year-round grazing system involving grazing of corn crop residues followed by grazing stockpiled grass legume forages was compared at the McNay Research Farm with that of the winter component of a minimal land system that maintained cows in drylot,. In the summer of 1995, two cuttings of hay were harvested from two 15-acre fields containing “Johnston” endophyte-free tall fescue and red clover, and two cuttings of hay were taken from two 15-acre fields of smooth bromegrass and red clover. Hay yields were 4,236 and 4,600 pounds of dry matter per acre for the tall fescue--red clover and smooth bromegrass--red clover. Following grain harvest four 7.5-acre fields containing corn crop residue were stocked with cows at midgestation at an allowance of 1.5 acres per cow. Forage yields at the initiation of corn crop grazing were 3,766pounds of dry matter per acre for corn crop residue, 1,748 pounds for tall fescue--red clover, and 1,.880 pounds for smooth bromegrass--red clover. Corn crop residues and stockpiled forages were grazed in a strip stocking system. For comparison, 20 cows were placed in two drylots simultaneously to the initiation of corn crop grazing where they remained throughout the winter and spring grazing seasons. Cows maintained in drylot or grazing corn crop residue and stockpiled forages were supplemented with hay as large round bales to maintain a body condition score of five. No seasonal differences in body weight and body condition were observed between grazing cows or cows maintained in drylot, but grazing cows required 87% and 84% less harvested hay than cows in drylot during the winter and spring respectively. Because less hay was needed to maintain grazing cows, an excess of 11,905 and 12,803 pounds of hay dry matter per cow remained in the year-round grazing system. During corn crop grazing, organic matter yield decreased at 27.3 pounds of organic matter per day from grazed areas of corn crop residue. Organic matter losses due to weathering were 9.4, 12.9, and 15.8 pounds per day in corn crop residue, tall fescue-red clover and smooth bromegrass-red clover. Organic matter losses from grazed and ungrazed areas during stockpiled grazing were 7.3 and 6.9 for tall fescue--red clover and 2.1, 2.9 for smooth bromegrass--red clover.

Effect of nitrogen supply on leaf appearance, leaf growth, leaf nitrogen economy and photosynthetic capacity in maize (Zea mays L.)

Leaf area growth and nitrogen concentration per unit leaf area, N-a (g m(-2) N) are two options plants can use to adapt to nitrogen limitation. Previous work indicated that potato (Solanum tuberosum L.) adapts the size of leaves to maintain Na and photosynthetic capacity per unit leaf area. This paper reports on the effect of N limitation on leaf area production and photosynthetic capacity in maize, a C4 cereal. Maize was grown in two experiments in pots in glasshouses with three (0.84-6.0 g N pot(-1)) and five rates (0.5-6.0 g pot(-1)) of N. Leaf tip and ligule appearance were monitored and final individual leaf area was determined. Changes with leaf age in leaf area, leaf N content and light-saturated photosynthetic capacity, P a,, were measured on two leaves per plant in each experiment. The final area of the largest leaf and total plant leaf area differed by 16 and 29% from the lowest to highest N supply, but leaf appearance rate and the duration of leaf expansion were unaffected. The N concentration of expanding leaves (N-a or %N in dry matter) differed by at least a factor 2 from the lowest to highest N supply. A hyperbolic function described the relation between P-max and N-a. The results confirm the 'maize strategy': leaf N content, photosynthetic capacity, and ultimately radiation use efficiency is more sensitive to nitrogen limitation than are leaf area expansion and light interception. The generality of the findings is discussed and it is suggested that at canopy level species showing the 'potato strategy' can be recognized from little effect of nitrogen supply on radiation use efficiency, while the reverse is true for species showing the 'maize strategy' for adaptation to N limitation. (c) 2004 Elsevier B.V. All rights reserved.

Summer dormancy in Festuca arundinacea Schreb.; the influence of season of sowing and a simulated mid-summer storm on two contrasting cultivars

Due to the shortage of information on summer dormancy in tall fescue (Festuca arundinacea, syn. Lolium arundinaceum), we tested the response of 2 cultivars of differing dormancy expression and growth stage to a range of summer moisture conditions, including full irrigation, drought, and a simulated mid-summer storm and analysed whether traits associated with summer dormancy conferred better survival under severe field drought. Autumn-sown reproductive and younger, spring-sown plants of 2 cultivars, claimed to exhibit contrasting summer dormancy, were established and then tested in summer 2002 under either long drought, drought+ simulated mid-summer storm, or full irrigation. The autumn-sown reproductive plants of cv. Flecha exhibited traits that can be associated with partial summer dormancy since under summer irrigation they reduced aerial growth significantly and exhibited earlier herbage senescence. Moreover, cv. Flecha used 35% less soil water over the first summer. However, the water status of leaf bases of young vegetative tillers of both cultivars was similar under irrigation and also throughout most of the drought (leaf potential and water content maintained over -4MPa and at approx. 1 g H2O/g DM, respectively). The summer-active cv. Demeter did not stop leaf elongation even in drought and produced twice as much biomass as Flecha under irrigation. Cultivar Demeter responded to the simulated storm with a decline in dehydrin expression in leaf bases, whereas no decline occurred in Flecha, presumably because it remained partially dormant. The younger, spring-sown swards of both cultivars had similar biomass production under summer irrigation but whereas Demeter regrew in response to the simulated storm, cv. Flecha did not, indicating that dormancy, although only partially expressed, was reinforced by summer drought. In all trials, cv. Flecha out-yielded Demeter in autumn regrowth. In particular, the severe drought in 2003 caused a 25% loss of the basal cover in cv. Demeter, whereas Flecha fully maintained its sward allowing it to produce a higher post-drought autumn yield. This work links summer dormancy with higher persistence over long, dry summers.