Data from: In situ measurements of animal morphological features; a non-invasive method

1) Measurements of morphological features are important for ecological studies, especially on free-ranging wild animal species. Conventionally, specimens either dead or in captivity are used for morphometric studies, which is difficult in the case of wild species for several reasons. Capturing would be even futile when research questions are relating to issues such as prey size selection or estimation of intake rate in under field conditions, where in situ morphometric measurements are inevitable. Remotely estimating morphometric features via, avian bill and head length are always laden with wide variances. Hence, we developed this technique for in situ measurements of animal morphology through digital photographs. 2) This non-invasive novel method for in situ estimation of individuals’ morphometric measurements uses digital photographs' metadata. We extracted the metadata (field of view, focal length, subject-distance etc.,) in Exchangeable Image File Format (EXIF) from digital images using freely available software, and the unknown sizes were estimated from each image. Cardboard (fish) models were, at first, used to estimate the accuracy of measurements comparing actual and estimated values, and the same protocol was repeated on museum specimens (bird). Finally, we checked the repeatability of this method under field conditions using free ranging Chickens (Gallus gallus domesticus). The results under different conditions were subjected to rigorous statistical testing. 3) Upon comparing the actual with estimated values, obtained from three distinct types of objects (1. Cardboard model, 2. Museum specimens, and 3. Live birds under field conditions), variations were found non-significant affirming the assumption that actual value (Y) = estimated value (X) with inconsequential variance of the Y from X. Moreover, both the relative and absolute errors between the actual and estimated values were so low, indicating that this method yielded neither over-estimation nor under-estimation. 4) The advantage of this non-invasive method is that it could generate reasonably accurate morphometric measurements at a relatively low cost for routine morphometric works in field conditions. We envisage significant utility of this non-invasive technique in different fields of science such as biology, morphology, ecology, evolution, wildlife, conservation science, agriculture, forensics and engineering design.

1) 形态特征测量在生态学研究中至关重要，尤其针对野外自由活动的野生动物物种。传统形态测量研究多采用死亡或圈养标本，但针对野生物种开展此类研究往往因多重限制难以实施。当研究问题聚焦于野外环境下的猎物大小选择、摄食率估算等议题时，捕获动物更是徒劳无功，此时原位（in situ）形态测量必不可少。通过鸟类喙长与头长远程估算形态特征的方法，往往伴随较大的方差误差。为此，我们开发了基于数码照片的动物形态原位测量技术。
2) 该非侵入性新型原位测量技术可获取个体的形态测量数据，其核心依托数码照片的元数据。我们通过免费可用的软件，从数码图像的可交换图像文件格式（EXIF）中提取视场、焦距、物距等元数据，并据此估算每张图像中未知的物体尺寸。首先以鱼类纸板模型为对象，通过比对实际尺寸与估算尺寸来评估测量精度；随后将相同实验流程应用于鸟类馆藏标本，验证方法适用性。最后，我们以散养家鸡（Gallus gallus domesticus）为研究对象，在野外环境下检验了该方法的重复性。不同实验条件下得到的结果均经过了严格的统计学检验。
3) 对三类不同测试对象（1. 纸板模型、2. 馆藏标本、3. 野外活体鸟类）的实际尺寸与估算尺寸进行比对后发现，二者间的差异无统计学意义，验证了"实际值（Y）=估算值（X）"的假设，且Y相对于X的方差极小。此外，实际值与估算值之间的相对误差和绝对误差均极低，表明该方法既未出现高估也未出现低估的情况。
4) 该非侵入性方法的优势在于，可在野外环境下以相对低廉的成本获取精度较为可靠的形态测量数据，满足常规形态测量工作的需求。我们预期该非侵入性技术可在生物学、形态学、生态学、进化生物学、野生动物学、保护科学、农学、法医学以及工程设计等多个科学领域得到广泛应用。