Weed population dynamics in a rain–fed maize field from the Valley of Mexico

Dinámica poblacional de malezas en un campo de maíz de temporal del Valle de México

Francisco Molina–Freaner¹ , Francisco Espinosa–García^2,* and José Sarukhán–Kermez³

¹ Departamento de Ecología de la Biodiversidad. Instituto de Ecología. Unidad Hermosillo. 83000. Hermosillo, Sonora. México. (freaner@servidor.unam.mx).

² Centro de Investigaciones en Ecosistemas. 58089. Xangari, Morelia, Michoacán. México e Instituto de Ecología. Departamento de Ecología Funcional. 04510. México, D. F. *Author for correspondence: (espinosa@oikos.unam.mx).

Received: June, 2007.
Aproved: February, 2008.

Abstract

Maize has a highly diverse weed flora in Mesoamerica, with a high proportion of native weed species. The long agricultural history of maize in Mesoamerica has imposed selection regimes that adapted native weeds to maize farming systems. These native weeds are expected to exhibit stable or increasing populations under traditional maize culture where herbicides are not used. In this paper, we study the population dynamics of major weeds occurring in a rain–fed maize field in the Valley of México. Eighteen species were recorded in the seed bank that showed a seasonal variation from 24 169 to 135 770 viable seeds m^–2. The four most abundant species represented 93% of the total seed bank and exhibited contrasting patterns of emergence that represented 2.2 to 3.1% of the initial seed bank. The first cohort was eliminated by mechanical cultivation while the second exhibited greater survivorship and fecundity than those emerging later. Weed species exhibited a wide spectrum of variation in finite rates of population increase (0.6–9.17): populations of Lopezia racemosa (λ=7.02), Galinsoga parviflora (λ=9.17) and Salvia tiliifolia (λ=1.81) exhibited values greater than one while Acalypha indica var. mexicana showed values less than one (λ=0.60). We argue that variation in population increase is associated with emergence patterns and with the maize monoculture and maize–alfalfa rotation system.

Key words: Acalypha indica, Galinsoga parviflora, Lopezia racemosa, seed bank, traditional farming systems, weed demography.

Resumen

Palabras clave: Acalypha indica, Galinsoga parviflora, Lopezia racemosa, banco de semillas, sistemas de cultivo tradicional, abundancia de malezas.

INTRODUCTION

Weed abundance in arable fields varies at different spatial and temporal scales, depending on agronomic and cropping system practices. Demographic studies have shown that herbicides (Bussan et al., 2000), crop rotation (Heggenstaller and Liebman, 2006) and crop spatial arrangement (Puricelli et al., 2002) affect finite population growth rate and consequently, weed abundance. Most of these studies have been conducted on modern industrialized agricultural systems where high inputs of agrochemicals and energy are employed. However, very few studies on weed population dynamics have been conducted on traditional farming systems.

The weed flora of Mesoamerica is highly diverse and has a high proportion of native weeds (Vibrans, 1998), a proportion which is greater in Mesoamerican agricultural areas than in other areas of México where agriculture is recent, possibly because of differences in agricultural history (Villaseñor and Espinosa–García, 2004). Presumably, the long agricultural history in Mesoamerica has selected native weeds that are well adapted to traditional farming systems that have resisted the invasion of alien weeds (Espinosa–García et al., 2004). Weeds are known to rapidly respond to selection in agricultural settings, exhibiting high synchronization of life–history traits to the crop cycle (Weining, 2005). Thus, the long history of Mesoamerican agriculture may have selected life–history traits that adapted native weeds to Mesoamerican crops. However, our knowledge about the natural history, life–history traits, and population dynamics of native weeds in traditional farming systems of Mesoamerica is poor.

Certain crops favor some weed species while constraining others (Lotz et al., 1991). Crop monoculture increases the abundance of certain weed species while crop rotation and diversified cropping systems reduce the abundance of others and increase weed species diversity (Swanton et al., 2006; Sosnoskie et al., 2006). Maize has a long history of cultivation in Mesoamerica (Harlan, 1992), and surveys of its weed communities have documented a large number of native species (Villegas, 1970; Vibrans, 1998). These weeds are presumably adapted to the maize traditional farming system, characterized by row sowing, frequent intercropping with beans, squash or both, and mechanical weed control applied before and 40 d after sowing. These native species are expected to exhibit stable populations or finite rates of population growth greater than one under traditional maize farming systems where herbicides are not used, and declining populations under other crops.

In this paper, we used demographic techniques to study the population dynamics of weeds growing in a traditionally managed maize farm in the Valley of México. Our objectives were to: a) document seasonal changes in the seed bank of major weeds occurring during the crop cycle; b) describe patterns of seedling emergence, plant survival and fecundity; c) estimate finite rates of population growth under a rain–fed maize cropping system.

MATERIAL AND METHODS

The agricultural system is located in the Valley of México, a closed basin in central México (Figure 1). Major features of soils, climate and major crops in the Valley have been described elsewhere (Rzedowski and Rzedowski, 2001; Villegas, 1979). Although irrigation played an important role in the development of Mesoamerican civilizations (Palerm and Wolf, 1972), currently most agriculture in the Valley is rain–fed (Villegas, 1979) and the most important rain–fed crop in the area is maize (Villegas, 1979).

We selected a maize field or milpa as our study system, close ( 2 km) to the Universidad Autónoma Chapingo in the State of México (Figure 1), located at 19° 30' N, 98° 53' W, at an elevation of 2250 m. Annual rainfall is 640 mm, mean annual temperature 15 °C, while local soils are Mollic inceptisols, derived mainly from lake sediments, with neutral pH and loamy texture (Enrique Solis–Villalpando, Personal communication). Locally, maize and alfalfa are the most important crops. Maize–alfalfa rotation is common, in which maize is rain–fed while alfalfa is irrigated with water from a communal well. During the year of our study (1983), the selected field had 7 years of continuous maize monoculture, after a period of 4 years under alfalfa. During 1983, the climatological station of Chapingo recorded an annual rainfall of 507 mm and a mean annual temperature of 17.9 °C (Figure 2b).

Rain fell occurred mainly between May and October (Figure 2b), and most agronomic practices were adapted to this rainfall regime (Figure 2c). The studied field was a small and rectangular (23 x 100 m) that was ploughed on April 5 and disked on April 30 (Figure 2c). Furrows were made on May 21 and maize was manually sown in May 30 on wet soil on the bottom of furrows, after the light rains of late May (Figure 2b). The distance between rows was 0.80 m and the mean density of maize was 56 900 plants ha^–1 . Maize growth in height and number of leaves exploit the period of the year where rainfall is concentrated (Figure 2a). The field was weeded twice during the early phase of maize growth by repeated mechanical cultivation with animal traction 40 and 47 d after planting in order to eliminate the first cohort of weeds and to hill the maize plants. Urea (200 kg ha^–1 ) was added manually around the plants between weedings. The maize dried in October, and plants were cut and cobs were harvested on November 20. The cycle ended with a disking operation that incorporated plant residues to the soil on December 15 (Figure 2c). There are no records of herbicide use for the control of weeds under this traditional farming system.

Seed bank

We set up a permanent plot (10X10 m) in the center of the studied field. The soil was sampled three times during the agricultural cycle (Figure 2c) using simple random sampling to document seasonal changes that occur in the seed bank. The first sample was taken on May 25 just after the furrows were established, before maize was planted and before weeds emerged (Figure 2c). This sampling represents the initial seed bank (N0) during the cycle and represents the source from which seedlings emerge. The second sample was taken on September 20, just before seed rain and after most seedling cohorts emerged (Figure 2c). This sampling (N1) attempts to record the number of seeds that occur after the losses represented by germination and before the input of new seeds. The third sample was taken on December 11, and represents the final number seeds (N2) after the weeds have dispersed their seeds, and just before the final disc operation that ends the crop cycle (Figure 2c).

The seed bank was sampled in the permanent plot using a soil probe (4.2 cm internal diameter). A preliminary analysis of the variability in the field indicated that the variance of the accumulated number of seeds of the most abundant species stabilized around the 25th sample. Thus, we decided to take 30 samples on each one of the sampling dates, representing an area of 415.6 cm2 . Given that the arable soil depth was on average 20 cm, we sampled the seed bank to a depth of 20 cm.

We buried 12 groups of 100 seeds of the four most abundant species at a depth of 10 cm in a contiguous undisturbed field in order to calculate the probability that a non–germinating seed has to survive to the next crop cycle. Seeds of each species were buried on July 14, 1983, in organza (fine mesh cloth) bags; three were recovered every two months until March 1984, and viability was determined through a tetrazolium test. A linear fit to data was used to predict the expected number of seeds surviving to the beginning of the next agricultural cycle in May 1984.

Emergence, survival and fecundity of weed cohorts

We installed 32 rectangular plots (15 x 62.5 cm) at random within the central field quadrat on June 11 to document the survival and fecundity of emerged cohorts of the most abundant species. These plots represented a total area of 3 m² and were distributed in pairs between the bottom and the crest of furrows. Every seedling that emerged within these plots was tagged with a numbered color ring. Plots were checked every week for surviving and new seedlings; seedlings emerging in different weeks were considered members of a different cohort. Seedlings were identified with an identification manual (Espinosa–García and Sarukhán, 1997).

Plant fecundity was measured toward the end of the life cycle of each species during October and November. We counted the number of reproductive structures (fruits or heads) on each tagged individual within plots. We also took samples of 100 reproductive structures of each studied species to estimate the number of seeds per reproductive structure. Plant fecundity was calculated as the product of the mean number of reproductive structures by the mean number of seeds (or achenes) per reproductive structure.

Finite rate of population growth

We estimated the finite rate of population growth using the expression proposed by Mortimer (1983) for an annual weed:

λ = Σ Ki Pi Fi + b, where: λ = finite rate of population growth; Ki = proportion of seeds that emerge from the seed bank in cohort i; Pi = probability of reaching reproductive stage by seedlings of cohort i; Fi = mean number of seeds produced by reproductive individual of cohort i; b = probability that the non–germinating seeds of the seed bank persist viable to the next agricultural cycle.

RESULTS AND DISCUSSION

Seed bank

The initial evaluation (N0) of the seed bank recorded 18 species of weeds with a total density of 53 913 viable seeds m^–2 in the 0–20 cm layer of the arable profile. The four most abundant species were Acalypha indica var. mexicana, Lopezia racemosa, Galinsoga parviflora and Salvia tiliifolia and represented 93% of the total seed bank, while Simsia amplexicaulis and Bidens odorata, the dominant weed species of the region, accounted for less than 2%.

These four species exhibited reductions in the seed bank from 44 to 72% during the period of emergence and cultivation, with respect to the initial number of seeds (N1/N0; Figure 3). The total number of viable seeds m^–2 recorded in the second sample (N1) was 24 169 while the number found in the third and final sample was 135 770. Lopezia racemosa, Galinsoga parviflora and Salvia tiliifolia exhibited an increase in the number of seeds while Acalypha indica var. mexicana showed a reduction in the seed bank (N2/ N0, Figure 3).

The abundance of weed seeds, their seasonal changes and the dominance of very few species in the seed bank in our study field fall within the range of variation found in other agricultural soils (Roberts, 1981; Forcella et al., 1992). Similarly, the number of weed species detected in the seed bank of our study field is similar to the species richness found in other agricultural soils (e.g. Dessaint et al., 1997). Thus, in quantitative terms, the seed bank of weeds growing in the traditionally managed maize field in the Valley of México appears to be similar to the seed bank of soils where technified agriculture is practiced. The major difference with respect to the seed bank of technified agricultural soils seems qualitative, as the floristic composition involves many native weeds (distributed in México and Central América or throughout the Americas; Vibrans, 1998) such as Acalypha indica var. mexicana, Lopezia racemosa, Galinsoga parviflora, Salvia tiliifolia, Amaranthus hybridus, Simsia amplexicaulis, Tinantia erecta and Ipomoea purpurea. Cosmopolitan weeds of European origin such as Sonchus oleraceus and Taraxacum officinale were found at very low densities in the maize field.

Linear regression was used to estimate the proportion of seeds surviving to the beginning of the next cycle, using the survivorship of seeds buried in bags (Figure 4). Using a regression model, the estimated proportion was 0.90 for A. indica var. mexicana (F=6.7; p<0.05; R²=0.40), 0.87 for S. tiliifolia (F=16.4; p<0.01; R²=0.62), 0.75 for L. racemosa (F=29.1; p<0.001, R²=0.74) and 0.74 for G. parviflora (F=20.7; p<0.01, R²=0.67).

Emergence, survival and fecundity of weed cohorts

Three to five cohorts were counted during the crop cycle (Figure 5). Weed species exhibited different patterns of weed emergence. A. indica var. mexicana showed a declining number of seedlings in successive cohorts while in the other three species the second cohort exhibited the largest number of seedlings (Figure 5). Emerged plants from all cohorts represented 2.2 to 3.1% of the initial (N0) seed bank (Figure 5). Cultivation invariably eliminated the first cohort that emerged in the field (data not shown). The second cohort emerged after weeding operations and their members had greater survivorship than cohorts emerging later (Table 1). Similarly, the second cohort produced a greater number of reproductive structures than cohorts emerging later (data not shown). In fact, for all species, the second cohort contributed with more than 90% of the reproductive structures produced by all cohorts: 91.9% for A. indica var. mexicana, 99.7% for L. racemosa, 99.5% for G. parviflora and 99.7% for S. tiliifolia.

Our results on the number of cohorts (3–5) recorded during the crop cycle and the percentage of emergence (2–3%) from the seed bank for the major weed species fall within the range observed in other agricultural soils (Roberts, 1981; Webster et al., 2003). The emergence of seedlings showed two contrasting patterns (Figure 5) in relation to weeding operations. Since the first cohort was eliminated and the second cohort had greatest success, the pattern of emergence had an important effect on the population size of major weeds as measured by the finite rate of growth (Figure 5; Table 1). At present, we do not know the mechanism responsible for the contrasting pattern of emergence but it is clear that the study of this mechanism could provide crucial evidence on the prevalence of native weeds under maize monoculture.

Therefore, our data show that survival and the probability of reaching the reproductive stage was greater in the cohort that emerged just after weeding operations (second cohort) than in successive cohorts. Similar evidence has been recorded in the study of the demography of weed species in other agricultural systems (Fernandez–Quintanilla et al., 1986; Puricelli et al., 2002).

Finite rate of population growth

The estimation of each of the parameters of the model proposed by Mortimer (1983) allowed determining population growth rates for each one of the studied species. The value of λ for A. indica var. mexicana was 0.60 while for L. racemosa, G. parviflora, and S. tiliifolia, values were 7.02, 9.17 and 1.81 (Table 1).

Our data on finite rates of population growth are similar to the range of values that have been recorded in weeds from diverse agricultural systems (Mortimer, 1983; Fernandez–Quintanilla et al., 1986, 2000). It is likely that the variation found among species in finite rates of growth is associated with maize monoculture and maize–alfalfa rotation. L. racemosa, G. parviflora and S. tiliifolia were the species that showed maximum emergence after weeding operations, exhibited λ > 1 and are typically abundant in maize (Villegas, 1970). In contrast, A. indica var. mexicana showed declining number of seedlings with successive cohorts, had the greatest number of plants eliminated by weeding operations, exhibited λ < 1 and is locally abundant under alfalfa (F. Molina–Freaner, Personal observation), where this species can survive the periodical alfalfa harvesting operations due to its basal branching habit and low height (Adame y Espinosa, 1992). Because in alfalfa there are no weeding operations until the first alfalfa harvest, it is expected that the emergence of Acalypha together with germinating alfalfa would result in greater seed set than delayed emergence. In alfalfa fields in the Valley of México, erect weeds are gradually eliminated by the periodical alfalfa cuts that occur before these plants reproduce (Adame y Espinosa, 1992).

Empirical (Westerman et al., 2005; Heggenstaller and Liebman, 2006) and theoretical (Jordan et al., 1995; Mertens et al., 2002) studies have shown that crop rotation imposes multiple stress and mortality factors that can contribute to effective weed suppression and deny the opportunities for growth or reproduction of continuous monoculture. Thus, it is likely that the different trajectories shown by major weeds in our maize field reflect differences in their ability to grow in maize or alfalfa. Our study field had seven years of maize monoculture after four years of alfalfa. It is likely that the native weed species that exhibited λ > 1 possess life history traits that allow a temporal escape to traditional weeding practices and produce sufficient seeds to increase in numbers. This set of native weeds may be well adapted to traditional management under the rain–fed maize of the Valley of México. Future studies on these native weeds should address the set of life–history traits that has allowed this group to persist so well under maize and the mechanism that has enabled them to resist replacement by weeds of European origin.

CONCLUSIONS

In this study we documented the demography of major weeds present in a rain–fed maize field in a maize–alfalfa rotation system. The seed density in the seed bank was similar to that in other agricultural soils. However, the floristic composition involved mainly native weeds.

The four predominant weed species represented 93% of the total seed bank and exhibited contrasting patterns of emergence and finite rates of population increase (λ ). The weed species typical in the maize phase, Lopezia racemosa (λ =7.02), Galinsoga parviflora (λ = 9.17) and Salvia tiliifolia (λ = 1.81) emerged predominantly after the first weeding operation, whereas Acalypha indica var. mexicana (λ =0.60), common in the alfalfa phase, reached an emergence peak before the first weeding operation.

ACKNOWLEDGEMENTS

We thank Heike Vibrans for critical comments on earlier versions of the manuscript, Dionisio Delgadillo for permission to study his farm, Enrique Solís Villalpando for information on soils from the area, and Rafael Díaz, Clara Tinoco and Rogelio Molina for field assistance.

LITERATURE CITED

Adame, C. J., y F. J. Espinosa G. 1992. Sucesión de arvenses en alfalfa (Medicago sativa). Memorias del XIII Congreso Nacional de la Ciencia de la Maleza . México. pp: 7.        [ Links ]

Bussan, A. J., C. M. Boerboom, and D. E. Stolenberg. 2000. Response of Setaria faberi demographic processes to herbicide rates. Weed Sci. 48: 445–453.        [ Links ]

Dessaint, F., R. Chadoeuf, and G. Barralis. 1997. Nine years' soil seed bank and weed vegetation relationships in an arable field without weed control. J. Appl. Ecol. 34: 123–130.        [ Links ]

Espinosa–García, F. J., y J. Sarukhán. 1997. Manual de Malezas del Valle de México. Universidad Nacional Autónoma de México y Fondo de Cultura Económica. México, D. F. 407 p.        [ Links ]

Espinosa–García, F. J., J. L. Villaseñor, and H. Vibrans. 2004. The rich generally get richer, but there are exceptions: Correlations between species richness of native plant species and alien weeds and in México. Diversity and Distributions 10: 399–407.        [ Links ]

Fernández–Quintanilla, C., L. Navarrete, J. L. G. Andujar, A. Fernández, and M. J. Sánchez. 1986. Seedling recruitment and age–specific survivorship and reproduction in populations of Avena sterilis spp. ludoviciana (Durieu) Nyman. J. Appl. Ecol. 23: 945–955.        [ Links ]

Fernández–Quintanilla, C., J. Barroso, J. Recasens, X. Sans, C. Torner, and M. J. Sánchez del Arco. 2000. Demography of Lolium rigidum in winter barley crops: analysis of recruitment, survival and reproduction. Weed Res. 40: 281–291.        [ Links ]

Forcella, F., R. G. Wilson, K. A. Renner, J. Dekker, R. G. Harvey, D. A. Alm, D. D. Buhler, and J. Cardina. 1992. Weed seedbanks of the U.S. corn belt: magnitude, variation, emergence and application Weed Sci. 40: 636–644.        [ Links ]

Harlan, J. R., 1992. Crops & Man. Second Edition. American Society of Agronomy and Crop Science Society of America. Madison, Wisconsin, USA. 295 p.        [ Links ]

Heggenstaller, A. H., and M. Liebman. 2006. Demography of Abutilon theophrasti and Setaria faberi in three crop rotation systems. Weed Res. 46: 138–151.        [ Links ]

Jordan, N., D. A. Mortensen, D. M. Prenszlow, and K. C. Cox. 1995. Simulation analysis of crop rotation effects on weed seedbanks. Am. J. Bot. 82: 390–398.        [ Links ]

Lotz, L. A. P., R. M. H. Groeneveld, B. Habekotte, and H. Van Oene. 1991. Reduction of growth and reproduction of Cyperus esculentus by specific crops. Weed Res. 31: 153–160.        [ Links ]

Mertens, S. K., F. Van den Bosch, and J. A. P. Heesterbeek. 2002. Weed populations and crop rotations: exploring dynamics of a structured periodic system. Ecol. Applications 12: 1125–1141.        [ Links ]

Moore, R. P. 1973. Tetrazolium staining for assessing seed quality. In: Heydecker, W. (ed). Seed Ecology. Butter worth Group, London. pp: 347–366.        [ Links ]

Mortimer, A. M. 1983. On weed demography. In: Fletcher, W. W. (ed). Recent Advances in Weed Research. Commonwealth Agricultural Bureau, Farnham Royal, U. K. pp: 3–40.        [ Links ]

Palerm, A. y E. Wolf. 1972. Agricultura y Civilización en Mesoamérica. Editorial SepSetenta–Diana, México D. F. 148 p.        [ Links ]

Puricelli, E., G. Orioli, and M. R. Sabbatini. 2002. Demography of Anoda cristata in wide and narrow–row soybean. Weed Res. 42: 456–463.        [ Links ]

Roberts, H. A. 1981. Seed banks in soils. Adv. Appl. Biol. 6: 1–55.        [ Links ]

Rzedowski, G. C., y J. Rzedowski. 2001. Flora fanerogámica del Valle de México. Instituto de Ecología y Comisión Nacional para el Conocimento y Uso de la Biodiversidad, Pátzcuaro, Michoacán, México. 1406 p.        [ Links ]

Sosnoskie, L. M., C. P. Herms, and J. Cardina. 2006. Weed seedbank community composition in a 35–year–old tillage and rotation experiment. Weed Sci. 54: 263–273.        [ Links ]

Standifer, L. C. 1980. A technique for estimating weed seed populations in cultivated soil. Weed Sci. 28: 134–138.        [ Links ]

Swanton, C. J., B. D. Booth, K. Chandler, D. R. Clements, and A. Shrestha. 2006. Management in a modified no–tillage corn–soybean–wheat rotation influence weed population and community dynamics. Weed Sci. 54: 47–58.        [ Links ]

Vibrans, H. 1998. Native maize field weed communities in south central Mexico. Weed Res. 38: 153–166.        [ Links ]

Villaseñor, J. L., and F. J. Espinosa–García. 2004. The alien flowering plants of Mexico. Diversity and Distributions 10: 113–123.        [ Links ]

Villegas D., M. 1970. Estudio florístico y ecológico de las plantas arvenses de la parte meridional de la Cuenca de México. Anales de la Escuela Nacional de Ciencias Biológicas, México. 18: 17–89.        [ Links ]

Villegas D., M. 1979. Malezas de la Cuenca de México. Publicación No. 5, Instituto de Ecología y Museo de Historia Natural de la Ciudad de México, México D. F. 137 p.        [ Links ]

Webster, T. M., J. Cardina, and A. D. White. 2003. Weed seed rain, soil seedbanks, and seedling recruitment in no–tillage crop rotations. Weed Sci. 51: 569–575.        [ Links ]

Weining, C. 2005. Rapid evolutionary responses to selection in heterogeneous environments among agricultural and nonagricultural weeds. Int. J. Plant Sci. 166: 641–647.        [ Links ]

Westerman, P. R., M. Liebman, F. D. Menalled, A. H. Hegenstaller, R. G. Hartzler, and P. Dixon. 2005. Are many little hammers effective? Velvetleaf (Abutilon theophrasti) population dynamics in two– and four–year crop rotation systems. Weed Sci. 53: 382–392.        [ Links ]

NOTA