{
    "document_id": 15918,
    "page": 68,
    "width_pt": 612,
    "height_pt": 792,
    "rotation": 0,
    "pipeline_version": "paddleocr-vl-1.6+paul-48",
    "engine": "paddleocr_vl",
    "source_run_id": 16030,
    "item_count": 2,
    "counts_by_type": {
        "footer": 1,
        "text": 1
    },
    "build_ms": 1,
    "items": [
        {
            "id": "p68.1",
            "order": 1,
            "type": "text",
            "bbox": {
                "l": 108,
                "t": 754.56,
                "r": 539.52,
                "b": 90.48
            },
            "text": "2004 study Trunk diameters were measured on March 3, 2004. 1ne number or maroon flowers on trees was counted every Monday, Wednesday, and Friday from April 12 until May 10. Twice during the flowering season, on April 28 and May 4, there were frost events and moist round hay bales were burned in an effort to prevent flower and cluster loss. Cluster counts were performed on June 7 and July 29. The number of fruit harvested was recorded every Monday, Wednesday, and Friday from August 23 to October 1. Methods for data collection were as described above.\n\nStatistical analyses Data were subjected to GLM analysis of variance and regression analysis using the statistical analysis program CoStat (CoHort Software, Monterey, Calif.). Treatment means were separated based on the least significant difference with a significance level of 0.05. Correlations were also tested between various means.\n\nResults\n\n2002 study Both vegetative and reproductive traits varied significantly among the six pawpaw populations evaluated (Table 5.2 and 5.3). TCSA among populations was significantly different, with the New York and Washington Co., Indiana accessing exhibiting more vigor than populations originally from Decatur Co., Indiana, and West Virginia. Total number of flowers between populations varied significantly, with trees from New York and Washington County, Indiana producing a larger number of flowers than accessions from West Virginia, Maryland, and Decatur County, Indiana. Length of flowering time also differed greatly, as populations from New York, Kentucky, and Washington County, Indiana blooming for a longer period of time than trees from West Virginia or Decatur County, Indiana. Flowering peak also differed significantly among populations, with trees from Kentucky and West Virginia having later flowering peaks than populations from Washington Co., Indiana, New York, or Maryland.\n\nPercentage fruit set was not significantly different between populations, with an average of 15% fruit set across all populations. Total number of clusters per tree did vary significantly, with accessions from Washington Co., Indiana having the greatest number of fruit clusters per tree. Number of fruit per cluster also varied significantly, with trees from Washington Co., Indiana, New York, and Maryland averaging more fruit per cluster than West Virginia. Trees from the West Virginia had the fewest number of fruit per tree.",
            "attrs": {
                "block_label": "abstract"
            }
        },
        {
            "id": "p68.2",
            "order": 2,
            "type": "footer",
            "bbox": {
                "l": 317.28,
                "t": 48.48,
                "r": 331.2,
                "b": 37.2
            },
            "text": "56",
            "attrs": {
                "block_label": "number"
            }
        }
    ],
    "excluded_furniture": {
        "first": [],
        "last": []
    }
}