Weed species in banana (Musa spp.) production systems in Türkiye

Abstract Background Banana ( Musa spp.) cultivation, has a centaury history in open fields in Mediterranean part of Türkiye has become enlarged for three decades due to benefits of covered production. Objective Weed species, their frequencies, distributions and reasons under current techniques and geographical attributes in banana fields and greenhouses were determined to provide data to develop strategies and systems for sustainable weed management and cultivation of banana. Methods Weeds and their coverage and densities were determined in banana fields and greenhouses surveying 2% of banana production acreage in 2021–2022. Results 68 different species of 25 families were recorded. The most common weed species were Cardamine occulta, Amaranthus retroflexus, Portulaca oleracea, Conyza canadensis , and Oxalis corniculata . The number of weeds, their coverage and densities in individual fields were not affected significantly by production method (open-field vs greenhouse) or banana cultivars but location. Years and field size affected the number of weed species while homogeneity and cycles affected weed coverage and density. Dominant species in greenhouses were C. occulta, Pilea microphylla, O. corniculata, Echinochloa crus-galli, Setaria verticillata , and A. retroflexus but Cyperus rotundus, S. verticillata, O. corniculata , and P. oleracea in open fields. C. occulta and Pilea microphylla were recorded first time in Türkiye. Conclusions The most common species and dominant species in fields are mainly aliens including both neophytes and archeophytes. Greenhouses and production techniques might have effect on weed problem. There are needs for further research on banana weeds including their controls to establish better strategies and control systems in banana management.


Introduction
Bananas (Musa spp.), originated from South Asia, are grown by 130 countries, which is led by India (Food and Agriculture Organization of the United Nations, 2023).They are an important produce in agriculture and economy of subtropical and tropical regions of the world.Banana, a cash crop in Türkiye, has produced along the coastal part of the Mediterranean region of Türkiye since the 1930s (Gök, Zaman, 2003).Production, starting on very limited fields has expanded year by year and reached 12,827 ha with 883,455 tons production per year (Turkish Statistical Institute, 2022).
Banana producing area of Türkiye is between 36 th and 37 th parallels which is the northest point of the banana production of the world.The area has typical Mediterranean climate that is rainy winters and drought summers.Especially temperature difference between day and night and between winter and summer as well as high summer and low winter temperatures is the main challenge in banana cultivation in the region.Open field banana production has almost centaury long history in a line between Mediterranean cost and Taurus mountains, which goes along central district and Anamur district of the Mersin province where foehn winds are common that makes banana growing possible.Banana production in covered environment has over three decades history in the region, which caused expanding banana areas via causing an increase in yield and shortening harvest duration (Gubbuk, Pekmezci, 2004).Covered production was mainly helped avoiding winter frosts that became detrimental from time to time, providing higher temperatures from 14 o C that not happen in late autumn and winter in the region and temperatures banana needs to start growing, and maintaining optimal humidity, 70-80%.On the other hand, under unproper conditions, temperature passes 34 o C that stress in plants start and diseases might be triggered together with higher humidity (Türkay, 2007).Heating is not a common practice; mist systems were used against frost.Drip irrigation and mini sprinklers are common irrigation methods (Sakarkaya, 2022).
Weeds are among the bottlenecks in banana production, which cause yield decrease and affect yield components as well as delay flowering and harvest (Santos et al., 2019).Especially young banana plants more prone to weed competition due to delay at maturity (Isaac et al., 2007).Yield components were negatively affected in weedy plots comparing to plots Nelsonia canescens (Lam.)Spreng were used as cover crop where weed density were reduced 60% (Fongod et al., 2010).Furthermore, there were more snails that bother the crop on upper parts of banana in weedy plots (Fongod et al., 2010).Commelina diffusa Burm., a weed in banana fields have been used as cover crop to prevent soil erosion; however, it increased nematodes and fungus that cause disease on banana (Isaac et al., 2007).
The studies have showed that many different factors can affect weed flora in banana cultivation from agricultural applications such as weed control and irrigation techniques to geographic, climatic, edafic and agronomic factors (Isaac et al., 2007;Fongod et al., 2010;Moura Filho et al., 2015;Cierjacks et al., 2016;Lanza et al., 2017;Ragas et al., 2019;Quintero-Pertúz et al., 2020;Wang et al., 2023).In order to develop strategies on researches and implementations on weeds in banana cultivation require to find out weed species and their phytosocioligical attributes.So far no a phytosociological study has been conducted in Türkiye except one where in a limited area in western Mediterranean part of Türkiye (Yılmaz et al., 2019).
The aim of current study is to find out weed species in banana cultivation under open field and covered systems in Eastern Mediterranean Region of Türkiye, and discuss possible research and implementation methods for a sustainable banana production.
The climate of area is typical Mediterranean climate with hot and dry summers and rainy and warm winters (Figure 1).The average temperature which was 17.3 o C for 1991-2020 period increased 0.6 o C and 0.9 o C in 2021and 2022, respectively (Turkish State Meteorological Service, 2021;2022).Many parts of the region have received more precipitation around 10-20% in 2021 and 2022 comparing to long term averages although drought in low levels was seen in some parts (Turkish State Meteorological Service, 2021;2022).
Banana fields were chosen randomly but proportional acording to district acreages.The area of field, production method (open field/covered), production cycles, and banana cultivar were recorded.In each field, five1m 2 areas were chosen as sampling points which were determined randomly.In each sampling point, plant species were recorded, their coverage (%) estimated, and individuals of each species were counted.The coverage (%) and density (individual in a squaremeter) of a species for a field were calculated as average of data in five sampling points and presented.Homogenity of weediness (%) for  ranked ones in the current study.Maybe larger study area and climatic and geographic conditions in Colombia caused higher number of weed species and families.In other studies, in narrower areas in Brazil, Poaceae and Fabaceae were also dominant families (Sarmento et al., 2015;Moura Filho et al., 2015).In the studies with different crops having similar durations with banana in the same region, 59 to 96 weed species belonging to 26 to 30 families were reported (Üremiş et al., 2013;Soylu et al., 2017;Özkil et al., 2019;Üremiş et al., 2020).
Less open field were visited due to the smaller number of fields and 17 species were recorded for coverage and 16 species for density for the highest two ranks.Species were recorded in three or more fields were POROL (8), SETVE (7), CYPRO (5), AMASP (3), DIGSA (3) and ERICA (3) for coverage, and CYPRO (6), SETVE (6), OXACO (4), POROL (4), DIGSA (3) and MEDSS (3) for density.The species, DIGSA and SETVE.with highest density or coverage in greenhouses are mainly neophyta except two grass species.It shows new and advanced technologies have driven invasion by plants in agriculture, which implies establishment of stronger quarantine measures and early warning systems.
Comparing to all fields, the median of the number of weed species in a field was 6 (1-15) but most of the fields had species between 4-7 (Figure 2).Mostly fields had a field were calculated weed occurred points out of five sampling points.Weed species were determined mainly using Flora of Turkey by Davis (1965Davis ( -1988) ) although online resources and newer literature were used as well.In addition, weeds were presented according to their families and cotyledon types.
Using the data, distribution of weed species were calculated for a given area such as total survey area, a district, a production method, or a province etc. and presented as frequency (%), which shows the percentage of fields that a species were determined regarding to given area.For instance: Where F g , F o or F g are frequency of an individual species as percentage in given area as all fields, open fields or greenhouses, respectively; N d , Nd or N dg are the number of all fields, open felds or greenhouses that an individual species was detected; N t , N to or N tg was the number of surveyed total fields, total open fields or total greenhouses.

Results and Discussion
In the study, 68 weed species from 25 families were determined in banana fields: Poaceae with 13 species followed by Amaranthaceae with 7 species, and Euphorbiaceae with 6 species (Table 2).In Colombia, 204 weed species from 54 families were reported with Poaceae as the most dominant family with 31 species followed by Fabaceae and Asteraceae (Quintero-Pertúz et al., 2020), which were also prominent families following to first three  up to 2 monocotyledon species although Poaceaae had the highest number of species in all survey area.Weed density per square meter was 8.80 individulas (median), which has reached to 37.20.The weed coverage mainly changed between 11.00% to 23.60% although there were fields covered with weeds up to 95.00%, means all field were weedy.
The weed parameters measured correlated in varying ways (Figure 3).The number of species correlated strongly with the number of dicotyledon species but less with monocotyledons.Weed coverage correlated with dicotyledons as expected.Weed density and weed coverage had high correlation as well.
The fields have two patterns in NMDS plot, but it was not possible to drove a concrete conclusion.Weed density and coverage had exactly same direction and power while the number of monocotyledon weeds did not show similar distribution with the number of dicotyledons or total weed species (Figure 4).Growing method for banana (open field vs greenhouse) or banana cultivars did not significantly affect any weed parameters measured but location especially districts level (Table 3; Figure 5).The median density in fields varied between 4.60 (Gazipaşa and Karaisalı districts) and 14.60 (Yumurtalık district).For weed coverage and the number of weed species changes were between 10.20 and 4.50 (Bozyazı) to 41.80 and 13.00 (Manavgat) (Data not presented).
AMARE was the most common species in two provinces, Adana and Hatay, CAROC in Mersin where has oldest banana fields and largest banana production area, and POROL in Antalya where has another largest banana area.The highest frequency of AMARE was recorded in Adana (63.16%) and Hatay (65.00%),CAROC in Mersin (58.82%), and POROL in Antalya (65.52%) (Table 4).
The other four factors have affected some parameters (Table 3).Year and field size significantly affected density and coverage while homogenity of weeds in field and growing cycles affected the number of species mainly (Table 2; Figure 6).
The crop cycles were grouped in three catagories.No-cycle (N) refers annually planted banana which is common in open field and applied in greenhouses as well; same banana root kept and used generally upto five years was denoted "S"; and root were renewed more than one times was showed as "R".Weed density and coverage were higher in greenhouses in R categories (Table 3; Figure 6).Only greenhouses with different crop cycling compared for the number of weed species.There were 62 species no-cycle (N) greenhouses or in the first year of its cycles but the number of species decreased to 37 in greenhouses applying renewed cycles system (R) (Table 5).

Conclusions
The weeds in open field are very similar to the other crops grown in the region but new species such as CARCO and PIMID have dominant in greenhouse banana production.There is no clue how these species have reached in the area although PIMID could be an escape from ornamental plant production or from houses.The muddy, warm and humid environment in banana greenhouses might be supporting increase and spread of  In greenhouse production banana roots were kept around five years (cycles) and renewed roots after 5 years have kept many years.This is a second study but more comprehensive than earlier one.Now, weed problems are clearer, which show more research needed to find out solutions.Labor based family farms rely on mechanical and chemical weed control, which should be improved.Manure use a common practice that needs to be applied carefully.These issues make farmer training necessary and find out farmers' profiles and their views on weed control to be able to draw a strategy and find out weed management systems.

Figure 5 -
Figure 5 -Effect of location (provinces and districts) on five weed parameters measured 5. Author's contributions HT, MÖ, İÜ, and AU: conceptualization of the manuscript and development of the methodology.HT and MÖ: data collection and curation.HT, MÖ, İÜ, SU and AU: data analysis.HT, MÖ, İÜ, and AU: data interpretation.HT and MÖ: funding acquisition and resources.HT and MÖ: project administration.İÜ and AU: supervision.HT and MÖ: writing the original draft of the manuscript.HT and AU: writing the first revised version.HT, MÖ, İÜ, and AU: writing, review and editing.All authors read and agreed to the published version of the manuscript.

Figure 6 -
Figure 6 -Effect of different factors on five weed parameters measured

Table 1 -
Banana production areas in the Mediterranean region of Türkiye (Turkish Statistical Institute, 2022), and sampled

Table 2 -
Weed species in banana fields in the Mediterranean region for Türkiye and their frequencies

Table 4 -
The most frequent speciesaccording to provinces provinces

Table 5 . The frequency of the weed species in banana greenhouses in the Mediterranean region Weed species Cycles and Frequency (%)
Authors Thank to Professor Necmi Aksoy to identifying some species and Professor Fatih Kahraman for sharing his knowledge on statistics.This study was financially supported by the Republic of Türkiye, Ministry of Agriculture and Forestry, General Directorate of Agricultural Research and Policies (TAGEM/ BSAD/B/21/A2/P1/2562).