Skip to content
Field Notes

The research side of permaculture.studio.

Agriculture & Soil5 min read

Berseem in Babylon: one field, one season, and a winner at the edge of the tested range

In a single-season Iraqi field trial, sowing Egyptian berseem on 5 November at 40 kg per hectare gave the highest fresh forage yield and crude protein.

What the research shows

  • In a single-season field trial in Babylon Province, Iraq, sowing Egyptian berseem on 5 November produced the highest green forage yield (18.53 t ha-1) and crude protein content (16.84%) among the four planting dates tested. Field trial
  • The highest seeding rate tested, 40 kg ha-1, produced the highest number of branches, green forage yield, crude protein and soluble carbohydrate among the three rates. Field trial
  • The second cutting produced the highest leaf-to-stem weight ratio (57.62%) and green forage yield (15.87 t ha-1) among the cutting stages. Field trial
  • Crude protein was highest at the second cutting (17.11%) but did not differ significantly from the first cutting (17.06%). Field trial
  • Soluble carbohydrate concentration was highest under the third planting date (43.45%) but did not differ significantly from the 5 November sowing (42.53%). Field trial
  • The authors state that the recommendations are specific to this location and season. Field trial

Each of these was checked against the abstracts of the papers it cites. See the receipts.

Berseem in Babylon: one field, one season, and a winner at the edge of the tested range

A field trial in Babylon Province, Iraq, found that Egyptian berseem (Trifolium alexandrinum L.) sown on 5 November at 40 kg per hectare gave the highest green forage yield and the highest crude protein content among the combinations tested [3]. That is the result. It comes from a single farmer's field, in a single winter season, with three replicate plots per treatment.

The authors say so themselves: their conclusion states that the recommendations are specific to that location and season [3]. The paper does not settle how to grow berseem. It shows which of three management levers moved yield and quality in one place, and roughly by how much, and hands anyone farming similar winter conditions a starting hypothesis rather than an answer.

What was actually measured

The experiment ran in the 2025–26 winter growing season in a farmer's field south of Babylon city. It was a factorial randomised complete block design with three replications, crossing four planting dates (20 September, 5 October, 20 October and 5 November), three seeding rates (24, 32 and 40 kg ha-1) and four cutting stages [3].

Recorded outcomes were branches per square metre, leaf-to-stem weight ratio, green forage yield, crude protein content and soluble carbohydrate concentration [3]. That is the whole quality panel: no fibre fraction, no digestibility measure, no mineral content, no intake or animal performance data. When the paper says quality, it means protein, soluble carbohydrate and the proportion of leaf to stem.

That last measure matters more than it looks. Leaf-to-stem ratio is the paper's proxy for how much of the harvested material is leaf. It is a weight ratio, not a nutritional outcome, and the abstract does not tie it to any feeding result.

The numbers

Planting on 5 November, the fourth date, gave the highest mean number of branches (427.4 m-2), leaf-to-stem weight ratio (54.24%), green forage yield (18.53 t ha-1) and crude protein content (16.84%) [3].

The highest seeding rate, 40 kg ha-1, gave the highest number of branches (256.2 m-2), green forage yield (15.68 t ha-1), crude protein (16.95%) and soluble carbohydrate (43.99%) [3].

The second cutting gave the highest number of branches (284.4 m-2), leaf-to-stem weight ratio (57.62%) and green forage yield (15.87 t ha-1) [3].

Two interactions are reported: the fourth planting date with the highest seeding rate produced the highest crude protein content (18.39%), and the fourth planting date with the third cutting produced the highest leaf-to-stem weight ratio (63.37%) [3].

What those numbers do not settle

Three things limit how far the ranking can be pushed.

First, the abstract reports means and calls them highest, but gives no p-values, no standard errors and no least significant difference for most comparisons. It flags non-significance twice: soluble carbohydrate was highest under the third planting date (43.45%) but did not differ significantly from the fourth (42.53%), and crude protein was highest at the second cutting (17.11%) but did not differ significantly from the first (17.06%) [3]. For the yield and protein gaps between planting dates and between seeding rates, the abstract does not say whether the differences were statistically distinguishable. From this text a reader cannot tell whether 18.53 t ha-1 sits meaningfully above the next best date or inside the noise.

Second, the winners sit at the edge of the tested range. 5 November was the latest planting date in the trial and 40 kg ha-1 the highest seeding rate. When the best treatment is the last one on the scale, the experiment cannot say whether the true optimum lies beyond it. Sowing later than 5 November, or above 40 kg ha-1, might do better or worse. The design is silent on that, and the paper does not claim otherwise.

Third, green forage yield is fresh weight — tonnes per hectare of green material, not dry matter. Fresh weight depends on how much water the plants held on the day of cutting; two treatments can weigh the same and deliver different amounts of feed. The abstract does not report dry matter yield, so the yield comparison is a comparison of wet material.

There is also a trade-off in the data. Soluble carbohydrate peaked at the third planting date, not the fourth, while yield and protein peaked at the fourth. The two carbohydrate values did not differ significantly, so this is a hint rather than a finding — but it is a reminder that the best date depends on which output you are optimising for.

The mechanism the study does not test

The paper measures responses. It does not explain them. Nothing in the abstract identifies why the November sowing outperformed the September and October sowings, or why the denser stand produced more branches and more protein.

The plausible agronomic story is about establishment conditions and canopy structure. Berseem is a winter forage legume, and sowing date sets the temperature and daylength the seedling meets. Seeding rate sets how many plants compete for light, water and nitrogen, which shapes how much stem versus leaf the canopy builds. Cutting stage sets how much regrowth the plant must fund from stored reserves. All of that is reasoning from how the crop is described, not something this trial demonstrated. The experiment was not designed to separate those pathways, and the abstract makes no attempt to.

Where the evidence runs out

The study is one site, one season, three replicates. The abstract does not report soil properties, previous crop, irrigation, fertiliser or weather for the trial period, so there is no way to judge how typical the field was or which of those factors might have driven the result.

The quality panel is narrow. Crude protein, soluble carbohydrate and leaf-to-stem ratio are not the same thing as feed value. Without fibre, digestibility or animal intake data, the paper cannot say whether the highest-protein treatment feeds better.

The statistical reporting in the abstract is incomplete. Rankings are stated without the error terms needed to judge them, so the differences between planting dates and between seeding rates should be treated as suggestive rather than established.

And the recommendation sits at the boundary of what was tested: the best date and the best rate were both the extremes of the design, which means the trial found the best option it was given, not necessarily the best option available.

What the paper supports is a local hypothesis worth testing on your own ground: for berseem in a Babylon-type winter, a November sowing at a dense rate is a reasonable starting point, and cutting stage shifts the leaf-to-stem balance. What it does not support is a general prescription. One field, one season, three replicates, and a winner at the edge of the range.

References

  1. [1]Partisan Banks and Creditor Coordination Within Loan SyndicatesRobert M. Bushman, Sang Woo Sohn, Jianxin (Donny) Zhao · 2026 · Journal of Accounting Researchdoi:10.1111/1475-679x.70097
  2. [2]Unlocking sustainable thermoelectric generation with zinc antimonideXinzhi Wu, Airan Li, Longquan Wang, Gang Wu, Jiankang Li, Zhao Hu · 2026 · Nature Sustainabilitydoi:10.1038/s41893-026-01986-6
  3. [3]Effects of planting dates, seeding rates and cutting stages on forage yield and quality of Egyptian berseem (Trifolium alexandrinum L.) in Babylon Province, IraqN M Hasan, A R M Nadhim, S O Hasanin, A A Ali · 2026 · Plant Science Todaydoi:10.14719/pst.17409
  4. [4]A quick appraisal of machine learning techniques in different aspects of rice cultivationAbhishek Singh, Nikhil Kumar Singh · 2026 · Discover Artificial Intelligencedoi:10.1007/s44163-026-02479-3
  5. [5]From race car telemetry to actionable insight: an open-source toolkit for researchers and practitionersNeila Danilevičiūtė, Dalia Mickevičienė, Uri Obolski, Gal Ziv · 2026 · Frontiers in Sports and Active Livingdoi:10.3389/fspor.2026.2005550
  6. [6]Monitoring-network change reshapes long-term Water Quality Index inference: a dual-frame analysis from GhanaMohammed Nurudeen, Addilat Abdul Mumin, Hannah Ayaba Tanye · 2026 · Water Sciencedoi:10.1007/s44533-026-00095-9
  7. [7]Consumer Preferences for Ice Cream Candy in a University Market: A Usage–Attitude–Image Analysis of Product Design and ValueJupert Jasser G. Abellana, Danya Nizreen Ampaso, John Dave V. Cafe, Tristan Troy Mikel Del Villar, Mariel T. Gabutan, Elena Maekaela Niña M. Patero · 2026 · TRANSEKONOMIKA: AKUNTANSI, BISNIS DAN KEUANGANdoi:10.55047/transekonomika.v6i5.1255
  8. [8]Foods and beverages reported in Gulf Cooperation Council countries: a scoping reviewGhadir Fallata, Amal A. Aloud · 2026 · Frontiers in Nutritiondoi:10.3389/fnut.2026.1928828
  9. [9]Plant bioluminescence for sustainable horticulture: Living biosensors, crop sentinels and precision crop monitoringParbin Huda Baruah · 2026 · Journal of Ecology and Conservationdoi:10.61577/jec.2026.1000014
  10. [10]Cotton (Gossypium barbadense L.) yield and fiber quality under nitrogen levels with topping practiceHouda A. Rabeh, Elsherbiny A. Elsherbiny · 2026 · BMC Plant Biologydoi:10.1186/s12870-026-10137-0
  11. [11]Agronomic Performance and Yield Dynamics of Wheat (Triticum aestivum L.) in Response to Variable Sowing Microclimates and Paddy Residue IncorporationAnkit Yadav, Raj Kumar Pal, Som Pal Singh, Sandeep Singh Sandhu · 2026 · International Journal of Environment and Climate Changedoi:10.9734/ijecc/2026/v16i115793
  12. [12]Pedogenic factors associated with soil fertility constraints in the Northern Guinea Savanna of Nigeria and their implications for soil managementYasin Agono Awwal, Sulaiman Lawan Ya’u, Lemuel Musa Maniyunda, Aisha Abdulkadir, Wasilatu Suleiman Ogala · 2026 · Frontiers in Soil Sciencedoi:10.3389/fsoil.2026.1991373
  13. [13]Experimental assessment of active and passive cooling configurations for photovoltaic modules under outdoor conditionsGleyzer Martins, Enio Pedone Bandarra Filho, Lucas Carvalho Martins, Yeyner Carrillo Pabon, Joao Rodrigo Andrade · 2026 · Frontiers in Thermal Engineeringdoi:10.3389/fther.2026.1970235
  14. [14]Physiological and yield responses of cassava (Manihot esculenta Crantz) to irrigationAl Imran Malik, Laothao Youabee, Saythong Oudthachit, Jochen Eberhard, Michael Scobie, Sarah Seton · 2026 · Scientific Reportsdoi:10.1038/s41598-026-80360-0

All 14 references resolved against Crossref on Nov 19, 2026. Full verification record.

Get the findings, not the noise

A short email when something worth knowing clears the verification gate. No takes. Unsubscribe in one click.